View Full Version : Weather forecasts for astronomers

2004-Jan-10, 09:40 PM
I find that general use weather forecasts aren't particularly useful as they don't tell me much about how good the sky is. But since starting flying, and having recently begun studying the meteorology for that purpose (it's part of the course), I've come across some new weather reports that are much more useful.

The MetOffice (http://www.metoffice.com/aviation/index.html) has a section for aviation in Britain.

There is obviously the synoptic charts for those in the know. For those not so, there are METFORM and AIRMET forecasts which I'm getting to know and will report on how to use those later, when I figure the out.

But, the thing I am getting to know are the TAFs and METARs, which are forecasts and actual reports respectively from individual airfields. Obviously, your best bet is to take the nearest airfield issue one to you and use that. TAFs and METARs are available at that link. Subscription is necessary but the basic stuff is free. I won't be posting any material from there, even out of date material since that will get me kicked out.

Fortunately, I don't believe knowledge of how to use TAFs and METARs are copyrighted material. So I will explain how to understand them, because they look weird at first and how they can be used for astronomy purposes.

First a METAR, which gives observed conditions at the time stated, ie not a forecast. Such a thing may appear like this (it's my own fake one):

METAR EGTR 092120 092120Z 23007G12KT 200V260 5000 -RADZ SCT020 BKN040 05/04 Q0999=

So to run through:
METAR EGTR This shows we're dealing with a METAR issued from the airfield with ICAO designation EGTR (Estree Aerodrome from which I do my PPL training, but they don't issue weather data so don't expect to see one from them).

092120 092120Z This is the time and date. The first two digits are the day of the month (the 9th), the next four are the Universal time. The data is repeated twice, with a Z for zulu on the second. (NB: the METARs given at the link above tend to omit one of the time blocks)

23007G12KT This is the wind. It's coming from 230 true (ie relative to the geographical pole and not the magnetic one), which is of no help to astronomers, and is travelling at 7kts with gusts at 12, which is of use because it tells us that we have fairly high wind and hence not very good seeing. Tonight is not a good night for that two hour shot of M33. If there are no significant gusts, the G12 will simply not be there.

200V260 This is the potential variation in wind direction that may not be noted if it isn't significant. This is of no use to astronomers.

5000 This is the visibility, which is what I really like, because it is the one thing that is really important yet never given in normal forecasts or reports. How often have weathermen said it was going to be a clear night, getting you all excited, but when the night came, while there were no clouds, the haze was awful? The visibility is given in metres and will give 9999 if visibility is greater than 10km. While this is specifically horizontal visibility, it is important, because it gives an idea of the haze and mist. If the number is poor, such as this 5000m, which is mediocre, then you know it won't be the best night for hunting those faint nebulae. If it 9999 and hence more than 10km, it shows that the sky is pretty clean and hence visibility will be good, once you get over the next bit.

-RADZ This is the weather. RA for rain with a - sign for light (no sign for medium and a + for heavy) with DZ for drizzle. There's a load of symbols, which I'll list later.

SCT020 BKN040 Most important of all: cloud data. Meteorologists measure cloud cover in oktas (eights of the sky). In METARs and TAFs, four designators are used. Few (FEW) is for 1-2 oktas. Scattered (SCT) is for 2-4 oktas. Broken (BKN) is for 5-7 oktas. Overcast (OVC) is for 8 oktas. The cloud base for the corresponding type of given is given as a flight level (ie hundreds of feet). So in this case, we have scattered cloud at 2000ft and broken cloud at 4000 ft. Obviously, we'd rather not have any. The flight levels may very well be irrelevant (unless you're considering heading up to a high very high mountain for an observing party). But, obviously knowing what kind of cloud we're dealing with it good.

05/04 This is very useful. The first is the air temperature. The second is the dew point. The first lets you know if you need a scarf. The second lets you know if you have to take precautions to make sure dew doesn't eat through your optics. If the dew point is close to the temperature, it means the air is very humid and hence there is a much greater likelihood of getting condensation on your lenses and mirrors.

Q0009 This is the QNH or atmospheric pressure at sea level for the airfield. It's not particularly useful to astronomers, unless you're a meteorologist and can make some conclusions from it.

Now for a TAF, which is a forecast for a nine hour period and occasionally for larger airfield, eighteen hour.

TAF EGTR 092120 092207 00000KT CAVOK 10/02 Q1013 TEMPO 0001 SCT020 BECMG 0607 OVC015=

092120 The time of the issue.

092207 The time of validity of the forecast. Valid on the 9th, from 2200 to 0700. The start and finish times are given as whole hours in the last four digits.

00000KT Wind is calm. Good seeing

CAVOK (pronounced ka vo kay) Ceiling and Visibility OK. Crystalline sky which is good, with no cloud below minimum sector altitude and no cumulonimbus at any height. Looks pretty good. This will replace visibility and cloud data if true. If not, then it'll be like the METAR.

TEMPO 0001 SCT020 Temporarily during the time, conditions may change to: between 0000 and 0100 to scattered cloud at 2000ft.

BECMG 0607 OVC015 The conditions will become at a certain time: between 0600 and 0700 to overcast at 1500ft.

Note: I think the above is right. 8-[ I'm sure someone like Waarthog will correct me on a few things and add a few other things. Either way, this tends to give much more explicit data on the conditions prevailing for a night of observation. Although there are many other sources of useful data of course.

2004-Jan-10, 10:06 PM

Yes, you are correct. However, in the U.S., METARs are slightly different. Not too different, but some of the parameters do not appear, and some do not appear the exact same way.

The following is how to decode a METAR issued by the Aviation Weather Center, one of NOAA's National Centers for Environmental Prediction:

10Z KLIT 200953Z 00000KT 4SM -RA BR OVC030 07/05 A3013 RMK AO2 SLP205

1. Time of Observation...in this case "10Z"
Note: This is not normally in METAR data, but has been added to simplify when the observation was issued. A routine (hourly) METAR will have a "Z" time...and a special METAR (taken in-between hours) will have an "SPC". To go from "Z" time to Central Daylight Time (CDT), subtract 5 hours (1000Z is 0500 or 500 am CDT). To go to Central Standard Time (CST), subtract 6 hours (1000Z is 0400 or 400 am CST).

2. Station...in this case "KLIT" or Little Rock National Airport.
3. Date/Time of Observation...in this case "200953Z"
Note: The first 2 numbers represent the day of the month, with the last 4 numbers representing the Z-time.

4. Wind Speed/Direction...in this case "00000KT"
Note: The first 3 numbers represent the wind direction. "360" is NORTH, "090" is EAST, "180" is SOUTH and "270" is WEST. The last 2 numbers represent the wind speed in knots. To go from knots to MPH, multiply by 1.15. "00000" means that the wind is CALM.

5. Visibility...in this case "4SM"
Note: The visibility is in "SM" or STATUTE MILES. If the visibility is 7 or higher, there are no visibility restrictions. If the visibility is less than 7 miles, something is restricting the visibility such as precipitation or an obscuration.
Visibility Restriction (Precipitation): Precipitation may come in several forms, with the most common including DZ (drizzle), RA (rain), SN (snow), IP (sleet), GS (small hail), GR (larger hail), or UP (unknown precipitation). The precipitation may have a descriptor such as TS (thunderstorm...usually combined with RA), SH (shower...usually combined with RA) and FZ (freezing...usually combined with RA or DZ). Precipitation intensities include - (light), no sign (moderate) and + (heavy).

Visibility Restriction (Obscuration): Obscurations, like precipitation, come in several forms, with the most common including BR (mist...greater than or equal to 5/8SM), FG (fog...less than 5/8SM), HZ (haze) and FU (smoke).

6. Sky Condition...in this case "OVC030"
Note: The first three letters indicate cloud coverage in eighths. 0/8 is SKC or CLR (clear), 1/8 and 2/8 is FEW (few clouds), 3/8 and 4/8 is SCT (scattered), 5/8 through 7/8 is BKN (broken) and 8/8 is OVC (overcast). The next three numbers indicate the height of the clouds in hundreds of feet. Example: "OVC030" means overcast at 3,000 feet.

7. Temperature/Dewpoint...in this case "07/05"
Note: The first two numbers represent the temperature in celsius. The second two numbers following the slant (/) represent the dewpoint in celsius. If the numbers happen to be preceded by an "M", that is a below zero celsius reading (example: "M05" means MINUS 5 celsius). To convert celsius to fahrenheit, or to determine the relative humidity, use a meteorological calculator.

8. Altimeter...in this case "A3013"
Note: The "A" stands for ALTIMETER with the next 4 numbers representing the pressure in inches (example: "3013" means 30.13 inches).

9. Remarks...in this case "AO2 SLP205"
Note: The remarks section of an observation may have a variety of data. The most common remarks include "AO2" indicating an ASOS observation and SLP indicating sea level pressure in millibars (example: "SLP205"means a sea level pressure of 1020.5 millibars).

Other aviation products, such as PIREPs (which are pilot reports provided during flight), are also encoded. The good news is, METARs, PIREPs, and Terminal Aerodrome Forecasts are all available decoded.

Additionally, for those wondering what a SIGMET and AIRMET are:

An AIRMET is an AIRman's METeorological advisory. It is mainly of specific interest to VFR (Visual Flight Rules) aviators. A SIGMET is a SIGnificant METeorological advisory. It is of interest to all aviators. There are different types of SIGMETs/AIRMETs:

Tango - Issued for turbulence.
Zulu - Issued for mountain obstruction and icing.
Sierra - Issued for IFR (Instrument Flight rules) condition.
Convective SIGMET - Issued for thunderstorms.

For more information on aviation meteorology, visit the Aviation Weather Center (AWC) at www.aviationweather.gov.

Happy flying!

Josh Gelman

2004-Jan-12, 02:46 PM
cleardarksky.com has a great chart for many different areas that plot out transparency, darkness, etc. They call it a "clear sky clock". I use it pretty much whenever I'm hoping to head out. Check it out. Wally

Hmm. the url didn't take. let me try this again. . .

Ok. I give up. y'all just cut/paste the link. (and someone tell me what I'm doing wrong please!)

trying 1 more time. Thinking it doesn't like being the first thing in the post. . .

nope. that wasn't it. OK, adding the http: stuff. . .


2004-Jan-12, 06:14 PM
I'm guessing it wants the http stuff. Also make sure the "Disable BBCode in this post" isn't checked.

Let me try...

cleardarksky.com (http://cleardarksky.com/)

That seems to work...

2004-Jan-13, 01:19 PM
I'm guessing it wants the http stuff. Also make sure the "Disable BBCode in this post" isn't checked.

Let me try...

cleardarksky.com (http://cleardarksky.com/)

That seems to work...

Well, as you can see, adding the http:// did the trick! Thanks.

2004-May-26, 05:54 AM
Free notification feature added:
"...The next generation Clear Sky Alarm Clock is open to subscribers. It will not start sending alerts until late May, but you can sign up now...email formats for every device from pager or cell phone to your desktop computer complete with the 30 hour forecast..."

2004-May-26, 10:12 AM
A few extra comments for those in the USA:

"CLR" or "CLR BLO 120" mean "clear skies below 12,000 feet." It is entirely possible (common, actually) to have thick overcast above that level. In TAFs (but not METARs), SKC means "clear sky."

Caution on TAFs: Most USA weather offices put a lower priority on TAFs than on significant weather (thunderstorms, winter weather, etc.), so it is possible that the TAF could be out-of-date (in terms of accuracy, I mean, since they're always updated at least once every 6 hours). Also, if there are scattered low clouds, a high overcast may be omitted for brevity (the FAA likes brevity). Along the same lines, emphasis in TAFs is always on low clouds.

Visibility is notoriously difficult to forecast, at least when it's variable (worse in some areas than others).

The National Weather Service now issues a graphical weather forecast (still considered "experimental," but it's done as if it were official). You can see the national map here (http://www.weather.gov/forecasts/graphical/sectors/conus.php) and click the map to zoom in on your region. To see sky cover, just point to the forecast time on the "Sky Cover" line. If you see good agreement across the area, you can be fairly sure the forecast is good. Again, though, if there are storms in the region, sky cover is just about the last thing on the forecasters' minds!

2004-May-26, 01:26 PM
The National Weather Service now issues a graphical weather forecast (still considered "experimental," but it's done as if it were official). You can see the national map here (http://www.weather.gov/forecasts/graphical/sectors/conus.php) and click the map to zoom in on your region.

That's cool! 8)

2004-May-26, 05:56 PM
Thanks for those descriptions! Everything makes so much more sense now!

I'm running gkrellm -- http://www.gkrellm.net/ -- and I have the gkWeather plugin installed, but I never could interpret the "ob:" output in the detailed display. This is very useful.

But a question for ya'll: why are the cloud heights in hundreds of feet, distances in miles and pressure in mmHG, but temperatures in C? Even in Britain? Can't we at least go metric in weather reporting?

Though I suppose this is related to the climate dataset problems described in Science a last week, where no-one will agree on when to sample, what units to use or how to report the results.... grumble... fething imperial system... grumble...

2004-May-26, 11:52 PM
But a question for ya'll: why are the cloud heights in hundreds of feet, distances in miles and pressure in mmHG, but temperatures in C? Even in Britain? Can't we at least go metric in weather reporting?

I forgot to mention that visibility in the USA is reported in statute miles (like 4SM), or as P6SM for "more than 6 miles.

The USA is extremely reluctant to use the metric system :( and our METARs and TAFs reflect that. We adopted the code about five years ago (?), leaving behind our "SA" observations ("Surface Airways") that were probably maddeningly hard to read for those schooled in METAR observations. However, the only metric units that were adopted were temperatures. (Since degrees C are less precise than degrees F, most USA METARs include a comment that looks like "T02090150" to specify a Celsius temperature that matches a Fahrenheit unit value. That's for records and such.)

I, personally, was shocked :o that we didnt' adopt visibility in meters, since they're relatively easy to convert to miles, if one should want to. Apparently, many pilots can't divide by 16 and move the decimal point a couple of positions. :D

I was similary surprised that we still put inches of mercury (only) for the altimeter setting.

Seems like our conversion to METAR was kind of half-hearted. I guess since the system reached maturity under the foot/mile/inch system, that's what we're stuck with....

Side note - the WSR-74 radar (from the early 1970s) was all metric, because the USA was planning to switch to the metric system. We didn't, obviously. When I started using it in 1989, it was a pain because we had to convert all the heights and distances to feet and miles. #-o

2004-May-30, 10:07 AM
Is there anything for australia :(