Wind-shear

[wlbragg]

How cool would it be to see this type of weather in FG, or maybe not?

I was organizing some hard disk space and came across this bit of video from a storm and some SLC we had in 2010.

https://i.imgur.com/Dn6ZT2t.mp4

[V12]

FG has windshear, You can observe it after METAR change when AW fails to interpolate condition from old wind speed, direction and baro pressure to the new one.

[Thorsten]

FG has windshear, You can observe it after METAR change when AW fails to interpolate condition from old wind speed, direction and baro pressure to the new one.

Well, AW never actually 'fails' to do that (this has been established in a couple of threads now) - unless you select a non-interpolating wind model of course.

Also, baro pressure has little to do with windshear...

[V12]

I many times described this effect. AW will always fail in this situation :
Live METAR AW mode,
ZULU time 0959
METAR station X from 0900 reports 20 kts wind from 270
then
METAR station X from 1000 reports 15 kts wind from 90
When AW fetch this new report, the weather engine fails to interpolate old weather into these new conditions and abruptly changes wind speed, direction and other parameters and my poor C172 crashed on the final approach.
I made many tests and many graphs about these changes. Now I use ActiveSky more than 3 years and NEVER seen similar behaviour. I performed many 3 or 4 hours flights over the Alps, Himalayas or other mountains over the all globe, but all weather changes was smooth. And on final approach I have many times heard aural warnings Windshear Ahead. Same warning was on ND too.

[Alant]

Wind shear is a local effect, and is reported in METAR as WS. e.g.https://www.anac.gov.br/en/safety/aeronautical-meteorology/conditions/wind-shear which reports wind shear on individual runways.
It is not the same as interpolating wind changes between two METAR reports.
Alan

[TheEagle]

If I understand it right, windshear is a rapid change in wind direction and / or speed, both of which can be simulated using advanced weather through the METAR string
- for changes in wind speed (gusts):

Code: Select all

10009G19KT

where 100 is the bearing of the wind, 09 the normal speed and G19 indicates gusts up to 19 knots.
- for changes in wind direction:

Code: Select all

060V130


after the above basic wind information, which indicates that the wind varies from 60 to 130 degrees.

[V12]

Wind shear is a local effect, and is reported in METAR as WS. e.g.https://www.anac.gov.br/en/safety/aeronautical-meteorology/conditions/wind-shear which reports wind shear on individual runways.
It is not the same as interpolating wind changes between two METAR reports.
Alan

You are correct if AW is working correctly and interpolating the weather.
But I observed many times and reported many cases, when the interpolation failed. And when interpolation will fail, You will obtain effect very similar to the wind shear. WS can be repetitive in short time period, interpolation fail can be repetitive too, but the frequency is same as the frequency of the METAR updates.

TheEagle :
Windshear is abrupt change of the wind speed and direction in the short distance, or small altitude difference. More info at https://en.wikipedia.org/wiki/Wind_shear
Letter G in the METAR report is acronym for gusty wind, not WS. String 060V130 means variable wind direction. Again, it is not wind shear.

[TheEagle]

ah okay … but you can get something close by increasing the turbulence setting, right ?

[V12]

Probably not. Turbulences are very different.

[Thorsten]

You are correct if AW is working correctly and interpolating the weather.

Several people have looked at the code and came to the conclusion that AW is correctly interpolating the weather. I've explained a dozen times that the result of correct interpolation is a sharp change if an interpolation point is fetched only long after the others should have expired, i.e. that the problem is with the input data, not with the interpolation - and AW has no control over the input data.

Yet, you somehow persist in claiming there is an interpolation problem in AW.

This is bad for a number of reasons, not the least that it distracts people's attention from the real issue and makes it harder to fix the real issue. If you go on like this, you might get a handful of people look at the AW interpolation routines. Like the last three who did, your next handful of helpers will also go through the routines, test them and come to the inevitable conclusion that interpolation is working correctly - and that the issue you describe doesn't exist.

So, apart from bad-mouthing my work (which seems to be a valid goal for you. given your post history) there's little constructive you actually accomplish by leading people astray that way.

***

Wind shear is a local effect,

Indeed, and it is so local we don't have the resolution in the model to resolve it - interpolation does changes on 20+ km scales (or a few thousand feet in altitude) and that is too large a grid to resolve actual windshear.

Gusty winds are the closest thing you can get, but they will be more random than real windshear. And of course a thermal represents a vertical shear layer.

Probably not. Turbulences are very different.

That, too, is misleading, because there is a close relationship between shear and turbulence - shear is a cause of turbulence, and turbulence advects and feeds shear.

[wlbragg]

I don't know that this video represents "typical" wind shear, although you could certainly see it for a bit and at a really low agl. It's more the point where rotation was starting to form because of such an unstable atmosphere. As Thorsten stated, wind shear is not wind suddenly changing direction at the same altitude, although I guess that might be possible. It is typically when wind changes direction sharply at different altitudes. Throw an updraft into the mix and that is typically what causes rotation to start to bend into a full on tornado. I've sat on runways out here waiting for this type of shear to subside before they will clear the pilot to takeoff. I have no idea if what I posted was strong enough to ground a heavy though.

Next storm day with strong shear I'll see if I can do a short time-lapse and demonstrate just how visible this effect can be if there is enough moisture in the air. Most of the type of shear that is so visible is usually prior to the supercell coming through but with a really high dewpoint or cloud cover at different levels. It can be clearly seen in those conditions.

[legoboyvdlp]

FG has windshear, You can observe it after METAR change when AW fails to interpolate condition from old wind speed, direction and baro pressure to the new one.

This has been discussed ad nauseum already, and I'm really not interested in going through it all again - but lets just be clear that AW does indeed interpolate.

I added a patch to FlightGear to improve its behaviour in one specific situation (weather points with radically different weather that were very close together), but in the other situation (weather points that are very far apart), a lack of data to feed into the algorithm does not mean that the weather is not interpolated once they are loaded.

[V12]

OK, if interpolation in AW works, then something other in AW is wrong, because only FG AW generates discontinuity and only in live METAR mode.
Again, more than 3 years I never seen any abrupt weather changes in P3D and ActiveSky. Freeware FSXWX / P3DWX interpolated weather between old and new METAR report from same WX station very similar to ActiveSky.

[S&J]

Bored bored bored bored bored.

My interpretation of what's been said is..... Yes the phenomenon does exist in FG but that's because of the data input.

Yes other programs use the same data and don't exhibit the phenomenon.

Conclusion: other programs when 'seeing' the data that'll cause this phenomenon have logic that handles it, reducing the effects to acceptable levels.

Going forward: isn't the aim to make logic that's close to real life and if the phenomenon doesn't meet that expectation. Well saying it's the Iinput data is just a copout.

[Parnikkapore]

Thus it should be considered if the interpolation should have a higher minimum transition time / distance so that it does not generate rapid (although smoothed) transitions that are most likely not really there.