Author Topic: Historical METAR Source  (Read 334 times)

xairbusdriver

  • Storm
  • *****
  • Posts: 1733
    • EW7115 (E7115)
    • KTNGERMA20
    • Mid-South Weather
  • Station Details: Davis VP2 wireless w/remote wireless Anemometer Kit
Historical METAR Source
« on: April 09, 2017, 09:01:23 PM »
I wanted some of the recent METARs from ATL for testing the 'remarks' section of a METAR module of my site. As slow as I am at ty pi n g, the current METAR had usually changed by he time I fixed any parsing problems. I did save many of them in a text file but they were not well organized by my "staff". [rolleyes2] [banghead] The interweb to the rescue! Here's a site <OGIMET> that has more METARs, TAFs, etc. than you could ever want! Enter the type of report, the starting and ending dates, and times and click the find/search button. The results are available as either html or plain text (my favorite!) to copy and paste into any text processor for cleaning up. Best news is that they are completely free, although numerous bytes were probably killed in the process...

WU has a similar capability but seems to supply only one day at a time. If you want the actual METAR or TAF, you have to click another button and the display will then be html with lots of extraneous text to clean out/up. There may be other sites, I just found the OGIMET site to be the easiest and simplest... always a high priority for my current brain! cmu:-)

xairbusdriver

  • Storm
  • *****
  • Posts: 1733
    • EW7115 (E7115)
    • KTNGERMA20
    • Mid-South Weather
  • Station Details: Davis VP2 wireless w/remote wireless Anemometer Kit
Re: Historical METAR Source
« Reply #1 on: April 12, 2017, 04:08:52 AM »
Many templates and other sites display METARs (and even TAFs). But all that I've seen simply display the text file basically as it comes from the NWS. That's fine for any one who reads them often. They are "built" from several dozen "standard" bits of text. OTOH, for those less familiar, usually non-aviators) 'decoding them can be more work than the viewer wants to exert. Besides, that's the kind of work we have computers for!

As an exercise in working/"learning" PHP, I decided to tackle converting the METAR text file into a more easily human readable display. Right now, it is a basic and very crude system, but it works better than 95% of the time, maybe more. The catch is that while most METAR reports are actually computer generated, they can have human entered data in the "remarks" section. Unfortunately, that means some humans make speling erers or they use the 'standard' terms in slightly different ways and combinations. That's where the real fun in the programming starts!

Here are some of the METAR reports I found for ATL around the hours when they were hit by severe thunderstorms; the ones that created such problems for Delta.

Note that the info on the thunderstorm ("TS") shows both an end ("E") and a begin ("B") time. Since the 'times' are only 2 digits, we know that they are in the hour before the report was generated. Below, you will see two more ways to display the beginning/ending of two different types of precip. [rolleyes2]
The other underlined text shows that it sometimes takes two text items to display something, miles of visibility, in this case. Without the space between the "1" and the "" things could get confusing.


This display mainly shows that I try to show the singular "mile" when appropriate. That comes from the way Apple goes to the trouble to show "second" for that split second in their  count-down displays. Why bother? Because some of us notice! [blush]
This image also shows a couple of the types of lightning. There are only four: IC, CC, CA, and CG. At first I only ever saw "LTGICCC" (Lightning Intracloud and Cloud to Cloud). Then the others started popping up. Usually they were combined about the way I listed them. Simple, just create some arrays filled with the 'usual' combinations: LTGIC, LTGICCC, LTGICCCCG... Wait! Here's one that did not have "IC" in it. And then I found one that had the order differently. I managed to work around it by simply 'deconstructing' the "LTG" string, but it would have been simpler if all the humans simply used the same format. [banghead]


Speaking of humans... what is the time that the observer wanted to report for when the "RA"in ended? What does "RAMM" mean. I suspect a computer automatically inserted the "RAMM" and expected the human to change the "MM" to a number for the minutes of the hour. Oh well, these folks are often rushed and probably bored/tired/sleepy.  [lol]


One last example of how these folks like to play fast and loose with precip reports. The first on reports the two types of precip ("RA", rain and "TS", thunderstorm) separately. It's a bit hard to tell, but html usually never breaks a word of text, the width of the display is usually allowed to control where a line of text breaks. In the first case, the line breaks with the 'space' between "RAB34" and "TSB18". Apparently, the observer sensed a different type of rain when the "RA" started at 34 but the "TS" seems to be continuing since 18!  [rolleyes2] Maybe he just forgot to write down when the TS actually ended? Who knows!

Below, he combined the two precip types, but that seems to be the only difference; the two times still seem backwards, at least. Either way, I just show them as I see em! In reality, I suspect one these 'two' reports somehow got duplicated at the archive site. The only difference is how the precip is coded. A corrected report should state that fact.


Right now, I have five airports hard-wired into the site. I think my next project may be to allow the viewer to request a particular airport and probably have that request saved so it will show on the next visit... that should keep me out-of-traffic for at least six months! cmu:-)

dfw_pilot

  • Gale
  • ****
  • Posts: 267
    • EW7118
    • KTXLEWIS26
    • WX Page
  • Station Details: Davis Pro2 Soil Temp/Moist
OGIMET
« Reply #2 on: April 14, 2017, 05:17:07 AM »
OGIMET - good find! Thanks!

dfw
A clear conscience is a great pillow.