left marrakesh this morning.
there was a fog cloud ONLY above the airport 
come on now⊠bring back full meteoblue as a toggle asobo.
i want a truly proper dynamic weather in FS2024 without these (forgive my dutch directness) garbage bubbles.
2 Likes
The current visibility in Marrakech is 8km and anything less than 10 triggers the âbubble.â
The question should be âwhatâs limiting the visibility in Marrakech alone?â There is a line of thunderstorms that just formed to the east, perhaps itâs blowing dust from the outflow? Or it could just be haze or smog. Temp/dewpoint is 33/19.
Or if itâs not just confined to that area in real life, then it becomes âwhy is limited visibility not depicted in the surrounding spaces (i.e. in between reporting stations)?â We know the sim doesnât do reduced visibility in the spaces in between, and thatâs what needs to be tackled.
cuz there is no station around it thatâs my whole point.
the closest station is gmmz which is quite a distance away.
it was literally a circular bubble JUST above the airport.
i keep seeing this all over the sim in my sessions.
itâs like the âforecast modelâ meteoblue shows nothing > the metar shows MVFR etc.
and you just get the metar drawing the data creating an effective bubble that blends nothing in from the meteoblue data.
which looks like a huge immersion killer when seen from above, looks highly unnatural.
1 Like
If they do historical weather well (which means an accurate representation of yesterdays (or any past day/date/time) weather conditions â which would ideally progress as the actual weather did), then it wouldnât matter to me if the current live weather wasnât great. But without well done historical weather, there is no good reason to not offer the option of the pure forecast model â none that Im aware of anyway â and they havenât said one meaningful thing about the weather in quite a long time.
I still donât know why Meteoblue hasnât released its own âforecasted weather addonâ. Iâd buy it. Presets are helpful but even FSXâs preset weather (or downloaded live weather) would morph and change over time. Start with clear sky and keep saving the flight and sooner or later youâd be flying in rainstorms. FS20 doesnât do that as far as I can tell.
Right now their plan is for just the past 24 hours, and it wonât be there at launch.
I think 24 hour old continuously updating weather would be perfection. It would give them enough time to fill in all the missing, unknown or incorrect data and combine with all the known good data to produce a really good weather simulation â truly as real as it could get⊠or at least the data would be.
1 Like
What I remember right after the sim was released was that the storms that I experienced in sim were pretty much right on what was actually going on at the airport. The first day I was flying out of KASH in the sim (a few miles from where my home is), and, just like what was actually going on, I was flying through passing cells of rain a mile or two each. Then they added lightning, which was neat at first, but then the sim weather reality went down from thereâŠ
When they introduced METARs, it got much worse, and now thereâs pretty much zero correlation of weather in the sim to real, at least in the Northeast US, and California/ Arizona where I flew in the sim the other day.
The other day, I flew from KPRB to KEDW and on to KFLG. The weather in the sim didnât match the METAR at all, including the ATIS did not match the METAR, winds were off by 10-15 knots in ATIS vs METAR. 20 miles out, ATIS claimed it was calm, METAR said 10 gusting to 15, which, for me, is not calm. 3 miles out, ATIS claimed 5 knots, METAR hadnât changed. Theyâve got a lot of work to do.
1 Like
Huh, I couldnât disagree more. The weather placement was awful and unpredictable when the sim first came out. Now itâs pretty darn close to being accurate. I do a 60-90 min weather briefing (amongst other things) each stream, I keep the VODs, and have plenty of evidence to back it up. In the last 10 days Iâve done five streams and the in-sim weather was right where I expected it to be each time and matched what AWC was putting out to within 30 minutes or so.
Maybe you were flying around some convective activity - that seems to be the remaining boondoggle as it is too momentary for the sim to capture it correctly. This includes outflow boundaries that can shift the winds.
When were you flying? Iâd like to take a look at the weather to see how stable it was, what was causing the gusts, etc.
It was a night time flight Iâd switched to daylight hours, but with Live weather. I trust that METAR was correct. It was the ATIS weather through ATC that didnât match the METAR winds. It was over the desert, so, so clear weather.
Maybe when you change the time, it uses a preset for the weather, even though I had Live weather active? But the winds did change as I got closer to the airport. Iâve seen that in the past, where the ATIS weather 20 or more miles out is completely different from the ATIS weather when you get close to the airport, as if they use your own local weather for the ATIS since they havenât loaded the airport, and its weather, yet.
And, yes, that early experience I related was convective weather, I was shocked at how âaccurateâ it was. When I say accurate, it wasnât that there was rain exactly where rain was, but, the experience of the passing showers in the general area (northwest Boston) was exactly what was going on, and the experience flying through them was exactly like Iâve done when Iâve flown on days like that, and it was pretty awesome. I havenât seen that detail since they started messing with the weather.
Personally, I think the main thing is, they need to find a way to much more smoothly blend the weather from Meteoblue and the METARS, which is I think what the OP really meant. Itâs definitely not an easy job to doâŠ
The visual presentation of convective weather was definitely better in the earlier versions of the sim. However, what I found was that it seemed to be just a general area of storms with wide geographical boundaries and they were often at least 6 hours old.
Many thunderstorms exist in smaller time slices than the sim can resolve. Other, longer-duration types (squall lines associated with fronts, long-track mesoscale convective systems) are there, but theyâre often 30 minutes or more behind reality. So the METAR change often precedes the arrival of the âweatherâ in the sim, and you get weird, anachronistic artifacts, which is a bummer.
So thunderstorms, specifically, have gotten both better and worse as the sim has evolved. But since theyâre the dominant weather type in the US from April through September and are the biggest hazard in aviation weather, itâs important that they continue to be improved.
I agree they need to better blend. But the problem is model-based weather diverges over time from the model run. So at some point the question becomes âwhy isnât Meteoblue showing the correct weatherâ versus why are we trusting the METAR. Thunderstorms, specifically initiation of, are a big phenomena that reveal the limitation of long and medium-range model runs. This is why we move to observations like radar, satellite, VADs, and surface observations once initiation has commenced.
METAR, of course, is subject to its own limitations - it is not the end-all solution, just one of a box of tools that can give a snapshot of reality as well as acting as a check on (and feedback into the subsequent) model runs.
A METAR bubble is an unintended consequence of a well-intentioned idea, so perhaps it can be solved programatically. The fundamental problem is, the real world doesnât always match the METAR either when you get in vicinity of the airport. Nature is by definition unpredictable, changeable and fast. I /want/ that type of real-world unpredictability in my sim. I do not want robotic machine exactness, that isnât realistic.
6 Likes
The problem with that, other than it being unrealistic, is thereâs nothing that gives us the ability to check whatâs happening ahead of us in real-time. Maybe the new weather presentation in FS2024 will give us some of those tools.
But I disagree that weather is wholly unpredictable. Certain phenomena are fairly predictable (once initiated) over a several-hour period. Is it predictable to the mile or minute? No. But it should be within several miles and/or several minutes. For instance, if I see a squall line bearing down on DFW from 50-100 miles away and have the movement direction and speed based on radar, I can make a pretty reasonable prediction of its arrival in the terminal area and the impact itâs going to have on traffic. That, versus using numerical models alone, which might miss that because back when they were run, they wouldnât know exactly where or when the storms initiate (then propagate forward).
With that in mind, Iâm not asking for robotic exactness. And METAR, which everybody gets hung up on, is often too slow, anyway. Iâm saying I know how weather generally behaves and Iâve never seen weather generation that isnât using some sort of close to real-time updates to a larger model get right the shape, movement, intensification, visual depiction, and representation in external products, all in three-dimensions. Pick maybe two of those, and thatâs not good enough.
What we have now, using models and a blend of updates, is the closest Iâve seen. But yes, they need to figure out some of the visibility issues and other gaps in the models. They need to tighten up the thunderstorm depiction. And most of all, they need to better blend the lighting and graphical representation. So in the end, I agree it could be better, but not for the same reasons generally claimed in these ongoing threads.
Iâm sure weâve all been stuck landing on a cross runway because the winds changed so fast they couldnât change runways fast enough. One in Nantucket clearly comes to mind for me. 
1 Like
Yes, or swirling winds in general - where the wind at the ASOS/AWOS on one end of the airport is not the same as the winds at the departure end of the runway, even with one-minute weather (which isnât used in the sim, anyway). Heck, SFO has this issue with ceilings!
1 Like
Are you responding to me? If you are, what I stated is most certainly not unrealistic. It is a factual truism. It is the definition of real-world events in flight, a non-matching METAR is something that happens daily.
As for the rest, you can debate with FinalLightNL or FlyingsCool some more, but I am out, I merely made a statement, I have no desire to go back n forth on something Iâve thought about deeply for the past several years â been on the METAR side for a while, been on the other for a while. This is why software toggle switches were invented.
Yes, modeled weather is unrealistic after so many hours. âUnpredictabilityâ is unrealistic to an extent. Conversely, so is ârobotic exactnessâ based on a single, hourly observation. But as I explained in detail, METAR are one of several observations necessary for mid-course, interim updates until the next numerical model is run and disseminated. Is there room to improve how itâs done? Sure. Start with fixing the surrounding visibility, then you wonât have the dreaded âbubbles.â
Itâs like METAR are put on a pedestal. The way folks latch on to the METAR in these discussions is like 1/10 of the issue and raises more questions that are rarely addressed. I donât think anybody who understands weather and wants a realistic solution wants them to be anything more than one aspect of feedback into the system. METAR is a simplistic strawman in many of these discussions.
1 Like
Once again, for those still interested, this is why models are both good and bad:
Hereâs a +6 hour model:
Hereâs what actually happened:
Note that the model is close, but it would have several terminals shut down compared to what actually happened at +6h. And pilots real and virtual would have no other way to know whatâs happening without some sort of real-time observation. Currently, the only ones we have are the ones afforded us by real aviation weather services, based on, you guessed it, real weather.
So, do you stick with the model and say âtoo bad, hereâs your âweatherâ anyway,â or do you refine and augment it with other observations?
Or do we just eschew the whole idea of observation and go with âunpredictableâ and thus unquantifiable in any meaningful way?
Iâll take what we have, warts and all, with the proviso that it has room for improvement rather than regression toward the nonsense it used to be.
2 Likes
Yeah for like the first year sometimes weather would be using 12 hour old data in some places.