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

[LDM #VST-440496]: NLDN timing

Hi Takeshi,

re:
> We are receiving the NLDN data via LDM at the Geophysical Institute,
> University of Alaska Fairbanks.
> 
> We are having an issue with interpreting the timing of the NLDN data.

OK.

re:
> We store the LDM data stream to files as it is. So, the files are in binary
> format. To interpret the binary data in a ascii format, I wrote a Python
> script following the data format description (
> http://www.unidata.ucar.edu/data/lightning/nldn.html)

Very good.  The information in this web page has been used by a wide variety
of folks to convert the binary data into formats for the particular package
that they want to use to display/analyze the data.

re:
> and a post on the support forum (
> https://www.unidata.ucar.edu/support/help/MailArchives/ldm-mcidas/msg00447.html).

I am the author of this reply.

re:
> The script works fine except interpreting the timing information. The
> timing of the data goes like this
> 
> ...
> 00:00:38.7
> 00:00:38.8
> 00:00:38.0
> 00:00:38.3
> ...
> 00:00:38.8
> 00:00:40.4
> 00:00:40.6
> ...
> 
> This implies to me that "00:00:38.0" after the first "00:00:38.8" should be
> "00:00:39.0".

Your assumption is that the reports are all in time order.  Given how the
data is collected, it is possible/likely that some reports from "earlier times"
will be received by the central collection system later than reports from "later
times".  As far as I know, Vaisala has never tried to time order the reports
so that the will come in sequentially.

re:
> I thoroughly checked my Python script but found no problem with it. As
> mentioned above, the script works just fine except the issue on the timing
> information.

Again, it seems that you are operating under the assumption that the
events reported should be time ordered; I do not believe that they are
now or ever were.

re:
> I attached a set of the NLDN data in the binary format. Could you dump the
> data in a text format and see if there is any issue in the timing of the
> data?

I don't think that there is any need to do this.  When I first wrote the
ldm-mcidas decoder to convert NLDN data into a format that is compatible
with McIDAS (MD file format), my examinations of the data showed that they
were not necessarily received in time-ordered sequence.  I made the decision
to not try to time order the data since the only way to do that would slow
down the ability to plot the data in near-realtime.

Please let me know if I have misinterpreted your question!

Cheers,

Tom
--
****************************************************************************
Unidata User Support                                    UCAR Unidata Program
(303) 497-8642                                                 P.O. Box 3000
address@hidden                                   Boulder, CO 80307
----------------------------------------------------------------------------
Unidata HomePage                       http://www.unidata.ucar.edu
****************************************************************************


Ticket Details
===================
Ticket ID: VST-440496
Department: Support Datastream
Priority: Normal
Status: Closed