Meteotemplate Forum

Lately, I have noticed that in certain weather conditions, (fog, rain, thunderstorms) the header backgound, (obtained from interactiveBanner.txt), does not match the current METAR report.
Here are 2 examples of this:

The METAR report that meteotemplate requested from http://tgftp.nws.noaa.gov/data/observat ... s/LPMR.TXT for my default METAR station (LPMR), was

2017/10/17 23:00
LPMR 172300Z 19010KT 9999 TS FEW007 SCT015 BKN030 16/15 Q1014

and the resulting InteractiveBanner.txt was: mostly

Note that InteractiveBanner.txt is not taking into account the TS (Thunderstorms) info of the METAR
--------------------------
Changing my meteotemplate default METAR (LPMR) to other nearby METAR (LPPT), the result was

2017/10/17 23:30
LPPT 172330Z 24009KT 9999 -RA FEW006 SCT020 FEW025CB 16/15 Q1015 RERA

and the resulting InteractiveBanner.txt was: partly

Once again, interactiveBanner.txt is not taking into account the RA (rain) info of the METAR
--------------------------

I have checked the header.php script and everything seems to be ok because all these possibilities are included Is there any reason why some of these conditions don't show up in my header background (and in interactiveBanner.txt)?

In Saratoga, the METAR and the current conditions icon, match perfectly (for my LPMR METAR)

Could be caching problem or it didnt load the METAR correctly at that time, the server it uses (NWS) is not very reliable.

Thanks Jachym.
I usully delete the cache before testing the METAR reports, but I agree with the possibility that in these particular situations the metar wasn't loading correctly. I will continue testing it from time to time and will let you know the results.

But one curiousity remains...
As I can see, the METAR report lenght and its content can vary a lot.
Yesterday it was a quite complex report, with TS (thunderstorms) info and different cloud covers at different heights (few clouds at 700ft, scattered at 1500ft and broken cloud layer at 3000 feet).
2017/10/17 23:00
LPMR 172300Z 19010KT 9999 TS FEW007 SCT015 BKN030 16/15 Q1014
and the resulting InteractiveBanner.txt was: mostly

At this moment the report is quite simple, just broken cloud layer at 2,000 feet
2017/10/18 10:00
LPMR 181000Z 16005KT 9999 BKN020 17/13 Q1013
and the resulting InteractiveBanner.txt is also: mostly

My question is: does the header background script take into account the TS info of the metar or does it just takes into account the cloud cover info?
Thanks.

It uses a general script for parsing METAR (available on GitHub), it should take it into account

Thank you so much Jachym!
No more questions... for now

I know this is sometimes problematic, it is also because METAR reports are not always 100% consistent and complying to the standard

Mine is doing the same thing right now.
Current condition are Windy, mostly cloudy. However the image in the banner is Lighning

Hi to all,
I'm coming back again with this topic as I would like to point out once more I think there is something does not is well running between interactive banner and metar/taf report.
I noted these inaccuracies only after the upgrade to version 15 black currant.

As you see in the attachments it's raining but the interactive banner shows blue sky...

I'm not very skilled neither about metar/taf nor about MeteoTemplate technicalities, all I did was cancel out the caches as suggested without success, but it's clear there's something wrong somewhere.

thanks
Norbert

Yes, but you must not forget this is cached, it must be, it is on every page and if it was updating the METAR with every page load it would slow down the page a lot.

Btw.... I never enabled the interactive banner in the DEMO because I prefer when it loads even faster