Meteoblue isn't a source of weather, in the sense that it's not an observation.  It is a computer model's guess of what the weather will look like 12 hours from now.  Before METAR integration for ceiling and visibility at airports, it was a common occurrence that the "real" weather in the sim didn't bare any resemblance to the weather that was occurring at that location in the real world.  This is because the Meteoblue model would often blow the timing of significant events like frontal passage by whole hours. It was entirely common to have an airport rendered as IFR when it should have been clear and a million, or vice versa.

This invalidated the use of any real world weather data for sim flying, and created issues on networks where everyone really ought to be experiencing the same conditions etc.  What we've got now works darn well though; clouds, vis and winds typically match real world metars quite well.

6 hours ago, Stearmandriver said:

If why live weather, it's so you can use real world flight planning tools like Jepp / Foreflight / WSI / etc. (Or even the wx reporting in Navigraph.) Also, so your weather matches everyone else's on a network.

This is why my own preferred solution would be for MSFS to use the Meteoblue forecast model throughout (rather than METAR synthesis), and for flight planning tools (e.g. Simbrief) to use the same Meteoblue forecast data. If Simbrief used a slightly different forecast (e.g. an older one, but predicting the same date/time) from what's in MSFS at that given moment, so much the better. In the real world, flight planning is done with forecasts that are often wrong, right?

(I've never used a flightsim network and almost certainly never will, so I admit I couldn't care less about whether an MSFS user on VATSIM is able to use the same weather as a P3D user. Given the very small percentage of MSFS users who are active on these networks, I never thought it made much sense for this tail to wag the dog, to be honest.)

I realize that a) MSFS is never going to go back to the pre-SU7 Meteoblue live weather and b) it may be technically or legally difficult, if not impossible, to build the Meteoblue forecast into Simbrief. But hey, I can dream, right?

James

2 minutes ago, honanhal said:

In the real world, flight planning is done with forecasts that are often wrong, right?

1.  Actual completed forecasts, whether they be TAFs or area forecasts or other products, aren't simply the output of a single computer model.  They are human-created by someone who considers the output of several different models and applies experience and judgement in local factors to produce the final product.  So they're wrong nowhere near as frequently or severely as a single model run like Meteoblue.

2.  Not only forecasts, but also the monitoring of current conditions are used for both flight planning and execution.  It's quite nice for real world aviators to be able to use real world tools in the sim.  That was impossible pre-su7.

12 minutes ago, Stearmandriver said:

1.  Actual completed forecasts, whether they be TAFs or area forecasts or other products, aren't simply the output of a single computer model.  They are human-created by someone who considers the output of several different models and applies experience and judgement in local factors to produce the final product.  So they're wrong nowhere near as frequently or severely as a single model run like Meteoblue.

2.  Not only forecasts, but also the monitoring of current conditions are used for both flight planning and execution.  It's quite nice for real world aviators to be able to use real world tools in the sim.  That was impossible pre-su7.

1. Fair enough, but this is still really a question of degree. Forecasts are forecasts, not 100% accurate in all cases, and that's why you build in tolerances to flight and fuel planning. Personally, I'm totally fine with having those forecasts be (even significantly) more inaccurate than real-world forecasts are for the purposes of simming and sim flight planning. (I know, won't anyone think of the safety of my virtual passengers? 😄)

2. I don't doubt that this is a quite nice-to-have in the sim for real-world aviators. That said, I'm not a real-world aviator, and neither are the overwhelming majority of MSFS users -- not even the majority of what we can shorthand as "hardcore simmers." So to me, and likely to most MSFS users, it doesn't matter.

I completely understand there are use cases -- e.g. "VATSIM user who wants to use real-world flight planning tools for MSFS" -- for which the post SU7 METAR synthesis is a big step forward. But there were huge tradeoffs involved in bringing that METAR synthesis to MSFS. It immediately massively downgraded the live weather for close to a year (remember the way practically every airport was blanketed in apocalyptic cumulonimbus rising from about 300 feet off the ground?). Even after they managed to make it significantly better over time, here's what we've still lost:

1) 90% of live weather clouds are still puffy cumulonimbus. Even larger cloud structures are often pieced together from these. (In the last few months we at least finally got back a decent overcast layer, but it's still not as sophisticated as pre-SU7.) Prior to the METAR synthesis, you had a huge variety of cloud types and shapes, especially stratiform ones. They looked extremely true to the real world (also quite beautiful in a lot of cases). Now, while they've improved since SU7, they often don't look quite right. That illusion/immersion has degraded. It frequently looks like a computer game, not real.

2) Weather isn't visible as a holistic system anymore. You don't get fronts and storm lines moving in the distance that you can actually see. Again, it's not that what we have is terrible on balance, but what was in at release was incredibly impressive, and it never came back post-SU7.

3) METARs create weirdness where the airport weather station doesn't match with the Meteoblue forecast around it. You get islands of weather (sometimes better, sometimes worse) around them. This is fundamentally unavoidable with a hybrid system, even with smoothing.

 

As far as I can tell, Asobo has done a pretty good job of trying to integrate METARs with Meteoblue. That's a hard thing to do well, and it's taken time, but they've massively improved on where we were when they started trying to combine them. But at the end of the day, for me, someone who ultimately doesn't care about having sim weather match current real-world METARs, the result is still worse than what we started with. We've lost the things I listed out above, and for gains I frankly don't value.

I understand why you and some other simmers prefer a METAR system, but I do wonder whether Asobo ultimately chose the needs of a vocal minority of users at the expense of a worse overall result for the majority.

James

(Sorry, that ended up being a long post, but the live weather system in MSFS is something I feel passionately about. As you can tell!)

43 minutes ago, honanhal said:

 

I understand why you and some other simmers prefer a METAR system, but I do wonder whether Asobo ultimately chose the needs of a vocal minority of users at the expense of a worse overall result for the majority.

It's this in a nutshell, the original Voting thread for METAR was harvested from the Vatsim crowd and Asobo bowed to pressure without thinking through and testing it thoroughly enough, here we are well over a year later still struggling with issues it's introduced. As others have rightly pointed out, it's worse in most cases than preSU7, the "bubbles" are still evident and a great immersion killer.

See, this is why I have to wonder if some of you just have an agenda. We see a prime example of the sim actually ignoring the innacurate METAR report to display weather that is representative of real world conditions (thanks to all of you who took the time to verify this) and yet you still see folks in here talking like something is fundamentally broken.

7 minutes ago, Krakin said:

See, this is why I have to wonder if some of you just have an agenda. We see a prime example of the sim actually ignoring the innacurate METAR report to display weather that is representative of real world conditions (thanks to all of you who took the time to verify this) and yet you still see folks in here talking like something is fundamentally broken.

some people just like to complain over and over again, shrugs..

have you not noiticed most forums nowadays have become pretty useless?

1 hour ago, MarcG said:

It's this in a nutshell, the original Voting thread for METAR was harvested from the Vatsim crowd and Asobo bowed to pressure without thinking through and testing it thoroughly enough, here we are well over a year later still struggling with issues it's introduced. As others have rightly pointed out, it's worse in most cases than preSU7, the "bubbles" are still evident and a great immersion killer.

I fly almost every day and absolutely perceive no bubbles or anything like that. Can you elaborate? For me a bubble would mean a sudden change of pressure or winds when crossing a certain point in space, or a sudden injection of clouds. As for pressure this hasn't happened for about a year, winds it does happen every once in a while BUT up high (so it's not down to Metars, they only influence surface winds), and a sudden injection of clouds I don't remember to have ever seen that.

16 minutes ago, Fiorentoni said:

I fly almost every day and absolutely perceive no bubbles or anything like that. Can you elaborate? For me a bubble would mean a sudden change of pressure or winds when crossing a certain point in space, or a sudden injection of clouds. As for pressure this hasn't happened for about a year, winds it does happen every once in a while BUT up high (so it's not down to Metars, they only influence surface winds), and a sudden injection of clouds I don't remember to have ever seen that.

https://forums.flightsimulator.com/t/nexrad-display-shows-metar-bubble-around-airports/

 

Just one thread example (loads more in other threads).

2 hours ago, honanhal said:

Prior to the METAR synthesis, you had a huge variety of cloud types and shapes, especially stratiform ones. They looked extremely true to the real world (also quite beautiful in a lot of cases).

QED:

 

The frustrating part is that Live Weather is considered "accurate" as long as there's CU, CS or TCU exactly where reported. The current Live weather does that admittedly better than the release version. But a warmfront has different cloud formations associated with it than a cold front, and their respective back sectors also look very different. Neither is depicted well in Live Weather these days. The OP is indeed a prime example that a METAR report has to be put into context. Only if the underlying meteorological conditions are simulated as well, you'll not be surprised to see solid OVC and light rain at your arrival eventhough the METAR reports NCD. The sim did it in this case for EGLL, but still lacks a lot nonetheless.

My point: the weather engine is capable enough, so why settle for a mediocre atmospheric simulation and not demand more? Sufficient accuracy AND plausibility AND immersion!

They do it for Scenery (WU) and Aircraft (AAU) and the core sim (SU). Doesn't the atmosphere- where the flying actually happens - deserve the same attention?

2 hours ago, Krakin said:

See, this is why I have to wonder if some of you just have an agenda.

[...] and yet you still see folks in here talking like something is fundamentally broken.

It's not an agenda, rather it's some people fixated on a particular bugbear and not fully understanding.

When METAR ≠ weather depiction in the sim, something is fundamentally broken, right? Well, as we can see in this case, no, it's not. However, the expectation leads to an inaccurate stance.

Another example is sound packs where "engine sounds are wrong in the cockpit." Expectation is for the engine sounds to be the same as in the cabin, where 99% of flight simmers usually sit in the real world, which informs their opinion. Whereas those that know recognise that sounds are different from the cabin.

4 hours ago, honanhal said:

I understand why you and some other simmers prefer a METAR system, but I do wonder whether Asobo ultimately chose the needs of a vocal minority of users at the expense of a worse overall result for the majority.

As always in life, the vocal minority wins.

That said, there is a need for live weather to closely resemble real reported conditions, since that's proper flight simming for the hard-core users.

I wonder whether an alternative system could co-exist based solely on Meteoblue forecasting, e.g. 'Live Weather no METAR', which would give a lot of people the pre-SU7 weather that they miss.

5 hours ago, honanhal said:

I completely understand there are use cases -- e.g. "VATSIM user who wants to use real-world flight planning tools for MSFS" -- for which the post SU7 METAR synthesis is a big step forward. But there were huge tradeoffs involved in bringing that METAR synthesis to MSFS. It immediately massively downgraded the live weather for close to a year (remember the way practically every airport was blanketed in apocalyptic cumulonimbus rising from about 300 feet off the ground?). Even after they managed to make it significantly better over time, here's what we've still lost:

I understand why you and some other simmers prefer a METAR system, but I do wonder whether Asobo ultimately chose the needs of a vocal minority of users at the expense of a worse overall result for the majority.

I think MSFS switched to a METAR system for live weather because the entire VATSIM, IVAO, Pilot Edge, etc, community needed live weather in MSFS to be consistent with the weather used for VATSIM, IVAO, Pilot Edge, etc.

Those users are a pretty big segment of MSFS users now.

46 minutes ago, abrams_tank said:

I think MSFS switched to a METAR system for live weather because the entire VATSIM, IVAO, Pilot Edge, etc, community needed live weather in MSFS to be consistent with the weather used for VATSIM, IVAO, Pilot Edge, etc.

Those users are a pretty big segment of MSFS users now.

IMO Live weather should be consistent with irl weather. If the METAR is wrong, the live weather shouldn't reflect that. That's just bad/late reporting and I imagine that occurs sometimes irl as-well which is why it helps to also read the TAF and crosscheck the METAR time with the current time. It's not iiregular for the METAR reported 30 minutes ago to be clear skies, then suddenly there's a downpour. MSFS should show the downpour, not a METAR report from 30 min ago.

6 hours ago, honanhal said:

2) Weather isn't visible as a holistic system anymore. You don't get fronts and storm lines moving in the distance that you can actually see. Again, it's not that what we have is terrible on balance, but what was in at release was incredibly impressive, and it never came back post-SU7.

What?? You most certainly do still get fronts and moving storm systems - which still come mainly from the MeteoBlue model. I did a flight last Sunday from Indianapolis to Fort Myers. The weather at IND was clear and sunny, and partly cloudy at RSW, but there was a narrow line of storms in the vicinity of Atlanta - and my FP route took me right over ATL. I could see the storm front clouds (in an east/west line) when about 100 miles north of ATL, and they got thicker and denser as I got closer. My only objection to the current system is that there is rarely any lightning anymore.

There used to be way too much lightning in LiveWeather, but they went overboard in suppressing it, in later SUs, and now it is rare to see it.

The clouds began to thin out to multiple broken layers and finally scattered as I flew south over the Florida panhandle and on down the west coast. In fact the position and distribution of the clouds was almost a perfect match for the r/w GOES satellite imagery for that area at that time.