There are so many unknowns that go into the user experience of weather and we’re really not collecting any useable information in these threads, just “I’ve seen it do this.” So we don’t even know if it’s a feature or a bug, as far as the intent of the development team.
If we really want to get to the bottom of it, we need to know the position of the aircraft, the real-world UTC time, what occurred (what was observed), what all the nearby in-sim METARs report (and whether the user is “within range” of any of those), and what real-world observations (surface, radar, satellite, or visible cameras) show at the same time and place. And not just a single data point or a small set from a single area.
I’ve done quite a bit of this research on my own, but haven’t really compared notes. I am, however, fairly satisfied with the results of the progress the sim has made over the past year or so. It was not even close three years ago, a bit better two years ago, and a year or so ago, it ended up close enough that I don’t even pay attention to the discrepancies anymore because microscale weather changes so rapidly that it’s inconsequential. Save for storm-scale convective activity, which needs more attention.
The non-precip visibility can still be a bit problematic, and I think we all agree the cloud rendering can be better.
I just don’t get it. I fly in the US also. I fly below 10,000 feet nearly all the time. I saw two or maybe three small thunderstorms with lightning last year and none for years before that. I pretty much never see any of the overcast conditions that I either see in real life near me or are reported in METARs. There are rarely any clouds above 15,000 feet at all. The clouds also do this very annoying thing where they tend to follow my altitude. If I’m flying at 7500 feet and the clouds are at my level and I drop to 2500 feet, the clouds will drop as well. Thats definitely not ‘clouds are where they should be’. It is not realistic - not in the slightest.
David explained this already, the forecast data we got before consisted of several day’s worth of weather. Any loop would have been at least that long and unnoticeable. Simply watch the many timelapses posted already from back then. It was behaving a lot closer to real world weather.
Except it simply did not. “Several days worth of weather” is not how real weather behaves, because chaotic perturbations on scales less than synoptic cause downstream deviation up to and including the synoptic scale, which all must be reconciled and fed back into the forecast model at some point. That reconciliation leads to the jumpiness encountered at any scale. The question is how much deviation and resulting reconciliation are we willing to put up with? What kind of scale? Hours, several hours, days, several days?
And what kind of tools will you use to plan and operate flights when the weather has deviated by that large of an amount?
Cirrus and thunderstorms are severely lacking. That has been discussed to death, and FWIW, I agree with you. I should have carved out an exclusion for those.
Thunderstorms actually suffer from the inverse of the problem we’re describing, because the weather generation engine doesn’t get granular enough to refresh them, unless they are part of a mesoscale convective system or mature squall line. The rest are too short-lived, oftentimes too small, and too dynamic to properly capture at-scale with the current regime.
Oh I’ve seen it snow too. But I’ll bet when you flew into the snow it hits the aircraft windows as rain. And also sounds like rain. What I had in FS9 with a 3rd party weather program was more realistic than what we have now in FS2020.
The clouds are what really lets down the sim in my view. They can look good for sure sometimes, but a lot of the time they are a mess. The colouring is wrong, they have poor colour banding, cauliflower, volcanic ash etc.
Wait, are you thus saying a 7 day forecast can not predict how real weather behaves? Remember the forecast was updated regularly every day and at most it was 6 hours old in the sim. This allowed live weather to behave realistically without jumps. You’re throwing around all these jargon to sound like the expert and asking where all the processing power will come from while meteoblue has been doing that all along until they moved to limited METAR data. What about all the bug reports showing the issues as clear as a circle around airports? Just ignoring that. You are happy that you can plan your flights now, whatever that means, while I that want to explore realistic live weather has to be satisfied with static presets or watered down live weather.
That’s correct, a forecast is a guess. What actual weather does is constantly feed back into weather propagation. Which feeds back into the forecast models, which get progressively shorter and more accurate.
7 days out is a guess as to what it will do. However, all the things that happen in the meantime make it so something different. If you run a system based on what a model thinks it’s going to do 7 days out, where it ends up will be very diverged from reality. Thus, it’s not how the weather actually is behaving.
False. I know how weather modeling and forecasting work. That folks can’t reconcile the discrepancies is not my problem, but I am trying to rectify that.
What “clear circle?” The radar display? The visibility? Addressed. The in-sim radar is nonsensical and I agree with that. The visibility is problematic, I agree with that. You’re not reading what I’ve said.
Because the in-sim radar is a related, but different problem. What cloud circles? I haven’t seen any in years. Are you talking about the visibility circles? Again, I agree that’s problematic.
Better than what is presented and what is being asked for here? Yes. But neither you nor I know exactly what data are being delivered to the sim and how the sim is interpreting and redisplaying them. What they have internally and present in their website is good. How it ends up in the sim due to the vagaries of the processing - your guess is as good as mine.
But for darn sure, a 6-hour cycle will get thunderstorms wrong. Thats a fact. You simply cannot know with the degree of accuracy necessary where thunderstorms (and their downstream effects) will occur until they initiate. Those storms, which are often discrete at the beginning of cycle often end up consolidating downstream and turning into an MCS or the frontal system itself. Once they’ve matured like that, they’re easier to predict. However, just as there’s diurnal heating, nighttime cooling will cause some to dissipate.
Again, 6 hours was/is not a close enough resolution. A thunderstorm that’s shutting down DFW, for example, may not shut down ADS, even though they’re only less than 10 miles apart. Or better, DFW and FWS (sorry, ACT was a stretch). Yet, a 6 hour forecast will not resolve that ands likely include both/all of them. That’s not good enough for flying.
Then you are either uninformed at best or disingenuous at worst.
And I don’t care if a thunderstorm arrives late or never show up based on the forecast, I can live with that. I simply want to continue my journey exploring realistic live weather like before. Right now that is not happening and I am waiting patiently for an update. But seeing this debate pop up again with no real change from all the other ones before makes me lose faith again.
Haha, you know what they say about anecdotes and data… Maybe I need to collect more anecdotes.
They also say ignorance is bliss.
Because it always, always devolves into groupthink based on anecdotes. There is very little evidence or hard data to gain a perspective on how off the weather really is or isn’t and why that’s the case. Heck, some of the “evidence” presented in the last hour or two is over a year old.
Right, you’re choosing ignorance. And honestly, that’s fine - I really don’t care. But you’re using your choice as a cudgel to purport that things should be a certain way. And I’m telling you the dots aren’t connecting and giving pretty solid rationale as to why.
And maybe slow down with projecting things onto me.
Your (and a few others in the group) understanding and observation of the behavior of live weather. But you just said that you don’t care if thunderstorms are accurate. Handwaving that, you might as well hand-waive what comes after.
And I have no clout or access to influence decisions like that. All I can do is educate, and of course, since this is all under the hood, I can only use the discrepancy between what is predicted and what is actually occurring as the focus. Because that’s the point the group keeps coming back to and it’s a flawed premise.
You want to feel good about what you see? I don’t decry that. I think the clouds, the blending, could all look better, too. But we head down this path every time, and I’m sitting here telling you all why it’s going to have poor results.
Why? Because you then could not plan your flight down to the exact METAR number? Why did I not have poor results for the first year of flying this sim until the “other” side chose accuracy of thunderstorm timings to the minute?
