Board index FlightGear Media

FG weather article on

Screenshots, videos, sound recording etc. taken in/with FlightGear.

FG weather article on

Postby Thorsten » Tue Apr 10, 2018 6:08 am

Dom Smith was kind enough to give us another opportunity to highlight FG's capabilities: ... FlightGear

It's an unholy amount of work to write these articles (I don't just write something, the text gets bounced back and force for readability), but it is really good publicity for us with O(10.000) readers per article. Feel free to have a good look and comment.
Posts: 11092
Joined: Mon Nov 02, 2009 8:33 am

Re: FG weather article on

Postby stuart » Tue Apr 10, 2018 9:55 am

Excellent article Thorsten! I think it does a fine job of explaining the complexity of the weather modeling. I particularly like the explanation of how the weather modeling impacts so many other aspects of the simulation.

I'll be interested to see the comments - I suspect this will highlight some gaps in other sim environments.

User avatar
Posts: 1469
Joined: Wed Nov 29, 2006 9:56 am
Location: Edinburgh
Callsign: G-MWLX

Re: FG weather article on

Postby CaptB » Tue Apr 10, 2018 12:27 pm

Great article, wonderful screenshot and excellent exposure for FG. Thank you for this!
Posts: 542
Joined: Thu May 23, 2013 6:36 pm
Callsign: EKCH_AP
Version: 2018.1
OS: Xubuntu, Win7 64

Re: FG weather article on

Postby legoboyvdlp » Tue Apr 10, 2018 7:34 pm

I passed this onto a friend of mine, who said it was overall good, but highlighted the following concerns,
it leaves out some important points on how terrain can hinder/affect the formation of convection clouds (They’re not called “convective” by standard)
Also, it’s surprising that it talks about glider pilots so much, but doesn’t mention one of the main ways of spotting thermals within the cloud layer

As far as I know, we do simulate those terrain effects?
User avatar
Posts: 7096
Joined: Sat Jul 26, 2014 1:28 am
Callsign: YV-LEGO
Version: 2018.3.1
OS: Windows 10 HP

Re: FG weather article on

Postby Thorsten » Wed Apr 11, 2018 5:16 am

I'm not sure what he expects - the article summarizes weather in FG, it doesn't go into all the gory details of thermal/terrain interaction and it also isn't an introduction how to find thermals for glider pilots.

As far as we simulate, the algorithm for generating thermals is fairly sophisticated, it uses (among other things)

* geographical latitude
* time of the day
* (user-specified) airmass stability
* terrain type
* terrain elevation relative to surroundings
* terrain elevation upwind

and I've verified for areas where I'm familiar with the cloud development patterns that it reproduces reality there (afternoon T-Storm development over Geneva between Jura and Alps was my best-studied test case). Doesn't mean it comes out like in reality everywhere - I don't know that for sure. Thermals in mountains are complicated.

If there's one main way of spotting thermals other than looking at active cap clouds, then it's been kept a secret from me when I learned glider flying - I can't comment further unless I know what he refers to.

Oh, and they are called 'convective clouds' in my meteorology book, and I just googled that NASA atmosphere science calls them that as well, so I gather the term is okay.
Posts: 11092
Joined: Mon Nov 02, 2009 8:33 am

Re: FG weather article on

Postby vnts » Sat May 26, 2018 3:30 am

Really nice article. article wrote:The second one on the other hand is a (crude) simulation of the state of the
atmosphere, driven by a few input parameters and randomization of what is unknown

This is a crucial point. METAR is so incredibly incomplete, vague, bounded & rounded off. The sources & accuracy are highly variable, for what info is covered at all (even visual observations can't be made if there's low visibility, or upper cloud layers are obscured). It's intended to aid safety. The automated & reviewed or human made observations are valid for one location at sometime in the past. Weather evolves and varies.

As complete information on atmosphere state in 3 dimensions+time, it's little more than 'Nice day out, isn't it?' or 'Weather has a bite to it today'.

So, what I'm getting at is, to end up with the desired weather or interesting weather pressing the 'Apply' button multiple times for any METAR should be the normal way to use FG with METARs (hard to deduce without UI telling?). It 're-rolls' on all the parameters METAR doesn't ever hope to cover. Hypothetically(??) PRNG seed could even be saved & reused to re-create spectacular weather exactly - for first-impressions & tutorials, or craft startup-states.


Picking a view and re-rolling unexposed parameters. Just a random shot messing about showing the very first grass iteration IIRC. Really like this shot because it shows rainbows need rain volumes, and rain actually falls under specific clouds in FG :D.

Simulated interactively...It's not a really lucky shot one may never see again using only live METARs because it's unlikely to be simming at the correct time of year, exact time of day, exact location to meet weather one feels like meeting: Rainbow 1st bow ==> 42 degree circle about point opposite sun (anti-solar point). Bow definition & intensity ==> droplet size (try heavy downpours). Sun position ==> time of day & year, & slowing time to catch lightning.

Combination of weather, rain on glass, lighting, terrain & OSM work is incredibly compelling. Old screenshots on January nightly near Bergen (autumn season). Was trying to capture new terrain & OSM features in one shot:

As an amusing side note; even some pretty rare cloud types like noctilucent high-altitude (100 km) clouds can be generated by the system.

Have to say, there were some interesting things in menu->debug/browse internal properties/ from what bits I've seen. IIRC there was a switch for noctilucent clouds. It's poetic to watch Auroras dance with time accel (e.g. 4x). Even from space. There was a control for aurora depth which changes colour - because of things like what excited atoms in atmosphere have how long to decay & emit before colliding due to density at different altitudes.

Etna (snowcapped in winter), while messing with volcanic activity & clouds:

Cloud layers & shadows on Heli&terrain (Hawaii):

It is awe--some that similar processes like convection play a part in phenomena as different as Auroras, Etna volcanism pictured above, and the storm cloud above. As I understand it: Auroras are caused by charged particles interacting and flowing guided by Earth's magnetic field lines contact the atmosphere. The magnetic field is caused by convective currents in the Earth's liquid core spinning by the Coriolis force due to Earth's rotation like cyclones do. Convection is also responsible for the circling conveyer belt in the solid but deformable mantle which transports colder surface rock down to the liquid part - thereby affecting convective currents & 'weather' in the liquid. Less dense, (partially) liquid magma rises up from below through the mantle like convective clouds created by rising air.

and if there is any open water, rain splashes can be seen impacting.

Even on puddles & aircraft surfaces.

Kind regards
Posts: 101
Joined: Thu Apr 02, 2015 12:29 am

Return to Media

Who is online

Users browsing this forum: No registered users and 2 guests