Thermals, Up & Downdrafts - More Realism Updates

In a german interview Jorg Neumann talks about Thermals and the weather physics in general.

Here are some of the keypoints he says:

  • currently they are working on high altitude turbulence
  • acknowledges that clouds are only visually represented, they have nothing to do with the ground or with rising air masses
  • Thermals are randomly distributed across the landscape
  • they were looking into the community feedback and definitely gonna revisit thermals and readjust
  • they want to beat condor in terms of weather and cloud physics simulation :astonished:
7 Likes

Apparently he doesn’t know that starting from SU 5, every cloud in his MSFS generates a thermal.
Apparently he does not know that starting from SU 10, different types of surfaces in his MSFS generate thermals depending on the type of terrine and the illumination by the sun.

1 Like

I think he meant that the additional new thermal system does not interact with the clouds (general updrafts generated by storm clouds was there before, but that’s something different).

That’s what’s happening, but there are also a lot more and bigger ‘‘thermals’’ randomly generated, because of the decision to use the ‘‘userfriendly’’ simplified thermal system instead of the more realsitic one (however that one would look like).

3 Likes

Now I’m totally confused, what about reports like Too small distances between thermals suitable for gliding, complaining about no thermals under the clouds but many elsewhere?

Depending on their density clouds can produce updrafts but wont form a classic thermal, that’s what’s happening since SU5, I don’t know if this was even intended or more like a bug, but you could soar under rain/storm clouds because of that.

The newly implemented thermal system however doesn’t correlate to the clouds in any form and thermals get spread out randomly (but still effected by other values, like temperature and so on).

The current system needs big adjustements / reworks, if they want to call it realistic.

Read this error message Errors of AMBIENT WIND Y dependencies on the values of flight conditions and terrain settings in SU11 and you will understand what termals now exist in MSFS and why many users complain about their excess and their shortcomings.

I’m sorry, just got back to FS2020 after they made the gliders/thermals in SU11, not familiar with the history, how exactly these are different from a “classic” thermal?

No problem, basically when you create an dark cloud layer you get constant updraft everywhere beneath that. That was discovered around SU5.This constant updraft is unrealistic and probably a bug.

A thermal has a core where the air goes up and on the sides you have down drafts / turbulence.

I can recommend the link Anri shared above where he explaines everything in detail. There you get an overview of what’s wrong with the thermal system currently.

2 Likes

Thanks, I did read it, also upvoted it, I understand there is an unrealistic updraft with dark clouds, guess I was confused because I thought this updraft exists for fair weather cumulus as well, which, even with the current limitations, wouldn’t be that bad, at least you could find lift under clouds.

1 Like

it’s actually a good question what happens under a flat dark cloud layer, basically looking like this:

thermals getting created by temperature differences between air and ground, hot air rises up and sometimes creates a cloud as it condensates, which is getting moved by the wind.

So when you have this big cloud coverage which got created by condensating water from the ocean/lakes and so on you don’t have any hot air rising up from the ground, because the sun is getting blocked by the clouds, which means no thermals and updrafts if I understand this correctly.

Maybe @ANRI8496 can correct me or explain it better, I would also like to know what really happens there :sweat_smile:

I hope they bring back the feedback snapshot, this topic is on top of the trends since the beginning and they really didn’t do any feedback snapshot since 3 months now or even longer… :upside_down_face:

5 Likes

20221230_121111
20221230_121124~2
20221230_121205
FYI these are the SOARING weather reference sources I’d consider fairly definitive. They’re full of valuable details based on probably at least 50 years of combined soaring experience. I’ll include two figures (“fair use”) related to thermals, one of which illustrates the ‘adiabatic lapse rate’ which explains the instability causing the air to rise in the first place and condense to form clouds, and another which illustrates the joint thermal/cloud life-cycle.


There is much more detail in the books providing a good foundation for why/how the air behaves this way but the basic idea is absolutely familiar to every RL glider pilot.

10 Likes

Now we’ve flown with the SU11 lift for a while, there are two specific issues which are so severe I’d like to highlight them as “bugs”:

(1) the “heated ground lift” is very strong and dependent upon ground type, with trees apparently generating SINK. Ridge lift is basically working (with a few limitations that would not be difficult to address) but the tree sink from the “heated ground lift” seems enough to overpower the ridge lift. This basically makes it impossible to create ridge soaring tasks anywhere there are trees such as the Eastern USA. In my RL experience the effect of trees on lift is extremely subtle (thermals extend later in the day a bit, and are a little moderated, no effect on ridge lift). Probably this effect of trees is a simple ‘bug’, instead of subtracting the full amount of the ‘sink’ due to trees a more reasonable approach would be to subtract maybe 10% of the otherwise existing lift.

(2) The “heated ground lift” is so widespread and goes so high that we’ve climbed under clouds and on ridges mistakenly believing it’s thermals or ridge lift we’re experiencing. MSFS glider pilots have learned to fly along the middle of valleys where the lift is strongest, rather than up against the windward slopes where the ridge lift should dominate. Somehow we need to be able to kill the unhelpful “heated ground” effect without disabling the other forms of lift.

These points aren’t intended to counter any of the various important soaring themes for which feedback has been provided in other posts. These points are more intended to highlight specific issues that could be checked during testing a weather update, because if these issues remain it’s clear the update has missed some fundamental concerns.

9 Likes

In the second figure, position 2 is depicted correctly. But, in the following positions 3, 4, 5… Questions: 1) Where has the wind disappeared? 2) How would the image change if the wind hadn’t disappeared? 3) Why did the heating of the earth’s surface stop at position 6?

Figure 2 is with a quite gentle wind, typical for a UK summer day, so there is no need to show more complex behaviour with the thermal initially remaining attached to the ground but pushed sideways an increasing amount as the wind increases with altitude and then the thermal detaching from the ground and drifting sideways with the cloud. It would be very wrong to GUESS the side profile of the thermal in this latter stage as it’s always more subtle than the school-grade first estimate would suggest, including additional warmed air from the original source joining the thermal, and no-one has ever accurately measured the side-profile of a thermal and its cloud drifting with the wind. The basic principle is originally the heated area of ground is a ‘fixed’ source so one end of the thermal is originally tied to that, but at some point the thermal detaches. Glider pilots do learn from experience that at least from about maybe halfway to cloudbase, to cloudbase, the thermal is more directly under the cloud than a school-room analysis might suggest. The subtlety is glider pilots under a Cu drift with the wind and use that Cu as their frame of reference (and stay under it) while if you’re nearer the ground you tend to have that as your frame of reference and definitely keep correcting upwind to stay in the thermal. These are details we’ve never discussed in the context of MSFS because they’re probably too complex and there would be a lot of guesswork from people that have never flown in one. For a FIRST APPROXIMATION it would be better to model thermals aligned vertically underneath the cloud, drifting with the cloud, than attempting something more complex and getting it wrong. Mostly glide pilots thermalling under Cu’s can (to a first approximation) turn in a simple circle and stay in the thermal, and the tracklog trace will show you drifting simply with the wind. It is FAR more common for glider pilots to thermal between 50% of cloudbase and cloudbase, than the bottom half of the thermal, so there’s simply more evidence there. A detail that glider pilots also know is if you ‘lose’ a thermal under a Cu (also known as ‘falling out of the thermal’) then on balance you are more likely to find it again by nudging cautiously upwind a little, which suggests a slight slope but not much. At the same time there is every possibility of another thermal joining the same cloud, I guess probably from the same source, and it’s not really possible when you’re flying to know if your seeing the behaviour of a ‘single’ thermal or multiple.

Re “why did the heating of the earth stop” - that’s more simple in principle and illustrated in Fig 2. In general the warmed air collects near the ground before it rises. I.e. the sun may heat quite a large area of ground and the adjacent air. At some point a local hotspot (probably) will cause local air to rise, that draws in other ground-level warmed air from nearby and the process accelerates with a relatively narrow column of warmer air rising up into the atmosphere typically eventually condensing to a Cu etc. You’ll see from the simplified skew-T diagram on the left of Fig 1 that on thermic days the air is actually unstable once it reaches a trigger temperature on the ground (but stable beneath that low-altitude ‘knee’ in the temperature curve - this is the primary reason why Asobo “heated ground lift” is fundamentally flawed). At some point the ‘pool’ of ground warmed air is exhausted (i.e. is replaced with cooler surrounding air) and the thermal continues to rise but is no longer fed from the ground. This is what looks like “heating of the earth’s surface stopped” but that isn’t the case, the sun keeps shining, the same surface will warm some more air and the cycle will restart.

NOTE details like “how large is the area of heated ground”, “what is the volume of warmed air”, “how long does the thermal remain attached to the ground”, “how long does the thermal remain usable for soaring” are NOT questions anyone has detailed answers for, it’s not as if there’s an army of scientists deploying dense arrays of sensors to measure this stuff. All we have to go on is the practical experience of pilots, particularly soaring pilots.

NOTE 2… it’s important to differentiate between “how does God create this complex behaviour” and “what code would I put in MSFS to simulate this behaviour”. It is not feasible to create realistic air movement from first principles at 60 FPS in a simulator. This is very true of thermals but also true of ridge lift. For the latter the oft-repeated mistake is to model a tiny patch of ground under the aircraft because that’s all you can afford with the generalised technique, and that results in what’s visible in all the retail sims that you can ridge soar tiny ridges buried downwind of much larger slopes that would be drowning them in sink in RL but the tiny local modelling in the sim doesn’t see that so it generates lift. The easy solution is to bias your ground sampling with a few samples much much much further upwind but sim devs rarely understand this because they are fixated on what they hope is a ‘generalised’ solution with a neat little box drawn around the user aircraft and one day computers will be faster and they’ll make the box a million times bigger and everything will be fine. There’s a similar analogy for thermals - modelling from first principles would be computationally infeasible but a simple cylinder of lift vertically underneath whichever cloud the pilot is under works pretty well for Condor and CumulusX (there are of course refinements, but that’s what they started with).

4 Likes

Forget not that it’s all relative, as during the day the general atmosphere warms individual heatspots won’t necessarily generate lift. Mid to late morning is usually a good time for thermalling, likewise Spring compared to Autumn.

Sim Update 12 in March, let’s hope for the mentioned ‘‘realistic’’ thermal version. :pray:

7 Likes

This bug report has over 180 upvotes. I hope Asobo won’t ignore our feedback.

The lack of communication from them about this topic is a little worrying.

There are several other reports from glider pilots going into great detail about the unrealistic aspects of the atmospheric modelling.

Please MSFS team give us a sign that you heard our feedback.

9 Likes

Yes, they didn’t update the feedback snapshot since 6 months or even longer. This topic would have been on the top of the list since the beginning, now it’s probably not even listed because it’s older than 3 months :unamused:

According to the roadmap there should have been a weekly update and if they bothered to post any feedback snapshot at all it was a repost from half a year ago! Did they fire the feedback manager or is this person on a half year vacation?

In the Forum itself you don’t get any feedback at all, if the topic is lucky you get a ‘‘feedback logged’’, but that also doesn’t mean anything. For example the ‘‘freelook freeze bug’’ is logged since Simupdate 5 and nothing happened.


But on a good note, Jorg Neumann stated in an interview that they are working on the thermals… But how and what they gonna improve, no idea.

I have the feeling they ignore the forum almost completely and just use feedback from some cry baby testers and that’s how we ended up with the arcade thermal version instead of the more realistic one which was an option.

6 Likes

Good news for Thermals, up & downdrafts in Sim Update 12 !

  • they picked up around 10 improvement subjects for Simupdate 12
  • better thermal height and alignment connection with clouds
  • rain gonna effect the thermals
  • better energy totals (there is as much air going up as it’s going down)
  • no more vertical wind limit, thermal strength can match real world numbers
  • further increased turbulence
  • option to manage turbulence: realistic, medium, low (doesn’t remove wind, gust or thermals)

(starting at 00:35:23)

13 Likes