SKEW-T questions

[Andrew Clope]

I'm still a bit unsure on Skew-T's and how they are obtained. My understanding is that they are just weather balloons that are released and then we are given a capture of the atmospheric conditions? I've been spending time on theweatherprediction and they were saying that it is an instantaneous capture of the atmosphere... How can this be? Doesn't it take time to travel 35,000 feet upwards?

Also, they said that they take 2 soundings a day (one at 00Z and one at 12Z). So do they just release 2 balloons a day, and does every weather office do this?

Thanks!

 

[Andrew Clope]

Update** If I'd have just read a little futher (lol) I would have read that some issues with the soundings are 1) that the wind blows the balloon downstream, so it's not a true vertical measurement and 2) the balloon takes time to travel to the Tropopause so it's not a "stillframe"

 

[Shawn Schuman]

Also, they said that they take 2 soundings a day (one at 00Z and one at 12Z). So do they just release 2 balloons a day, and does every weather office do this?

Thanks!

Most offices generally do two soundings a day, though sometimes there are special soundings at 06z or 18z or other assorted times. You can get an idea of that by checking the SPC sounding archive. You'll see most offices do 00z/12z and then one or maybe a couple offices will do soundings at other times.

http://www.spc.noaa.gov/exper/soundings/

 

[John Moore]

Skew-T's are also produced by models. That means you can get a forecast Skew-T. Importantly, you can get an estimated Skew-T for "now" for a spot where there is no sounding, or at a time when there is no sounding.

Models use more than radiosonde soundings (weather balloons). They also get input from commercial aircraft sensors, from GPS measurements of moisture, from satellite "sounders" (radiometers), and probably other sources.

A good site for model based Skew-T's is College of DuPage's very nice, recently revamped model site. At this site, you pull up a model map for a given time, sector and product, and then click on the spot where you want a Skew-T.

 

[Jeff Duda]

... that it is an instantaneous capture of the atmosphere... How can this be? Doesn't it take time to travel 35,000 feet upwards?

Yes, that is a known limitation of using balloons to obtain profiles of the atmosphere. However, at least on the larger scales (synoptic and above, and probably on the meso-alpha even), atmospheric changes on the length of time the balloon spends gathering data are too small to be of importance. So soundings obtained through this method are assumed to represent an instantaneously obtained vertical profile.

If I'd have just read a little futher (lol) I would have read that some issues with the soundings are 1) that the wind blows the balloon downstream, so it's not a true vertical measurement.

Also true. Again, however, the atmosphere does not change significantly even with the horizontal displacement of the balloon during ascent, which generally is no more than 100 or so km. While it is wise to make adjustments for these inadequacies when performing research, for operational forecasting purposes the error is generally too small to worry about. There are exceptions, but it is generally impossible to correct for sonde drift anyway.

 

[Andrew Clope]

Thanks for the responses!

Maybe a dumb question, but I assume that they recover these balloons after their travels right?

 

[Jeff Duda]

I assume that they recover these balloons after their travels right?

The technology exists to recover them, and I believe the instrumentation is packed with instructions on how to return it to an NWSFO if found, but I think the statistics on recovered instrumentation is pretty low. I think most of the time the package falls into a rural area where no one comes across it. The offices do not actively attempt to search for them.

 

[Mike Johnston]

Thanks for the responses!

Maybe a dumb question, but I assume that they recover these balloons after their travels right?

The balloon itself eventually disintegrates because of the very low atmospheric pressures encountered at extremely high altitudes. Most of the time, the instrumentation is lost, but occasionally you hear about them being recovered and turned back in to the NWS, etc.

Now, how much data is actually gathered from aircraft and used in models, I've often wondered about myself. In theory, since there are thousands of flights a day, you would think it would be significant. But, in practice, I'm not sure how dense these measurements are actually taken.

 

[Tim Supinie]

Now, how much data is actually gathered from aircraft and used in models, I've often wondered about myself. In theory, since there are thousands of flights a day, you would think it would be significant. But, in practice, I'm not sure how dense these measurements are actually taken.

Quite a bit, actually. Mostly temperature, pressure, and winds, from what I understand, but dewpoint sensors are on some newer aircraft. (Turns out, reliably measuring atmospheric moisture content is hard.) That's nice for flight-level data while en route, which is incredibly useful in a numerical model. However, you only get profiles that could be plotted on a Skew-T near the hubs, when the aircraft are taking off and landing, and you probably already have a balloon sounding from close by, anyway.

As for the original question, I feel the need to make the distinction between the Skew-T (the full name is "Skew-T log-p") and the data that are plotted on it. The Skew-T is just a type of diagram for upper air data that has a few nice thermodynamic properties and is (relatively) easy to read. Also, as mentioned before, there's nothing that says the data that are plotted on a Skew-T have to come from a weather balloon. Technically, any data that varies in temperature and pressure can be plotted on a Skew-T, which can be either from a balloon or from a model. These data could also be plotted on a number of different diagrams like a Stuve plot, or a Tephigram. You probably don't need to worry about those, though, as the US seems to be pretty set on the Skew-T.