From what I’m seeing post-1.15.10, it seems like the Live weather went from sometimes pulling a forecast too far into the future to always pulling (warning: speculation) what appears to be an outdated forecast snapshot. 15.10 definitely changed something, but it’s still not lining up with reality.
Comparing with the Meteoblue website, the cloud and precip positions on the world map look out of date by at least 12-24 hours. Fronts in the US generally move west to east, and their positions in the sim are west of their current real-world positions.
In-game conditions appear to be out of date in the same way displayed on the world map. It’s raining at Chicago Midway airport IRL right now (13:00Z), but it’s mostly sunny when I load into the game.
Oddly, when I select Midway as my departure point, the little weather picture preview in the top right shows rainy conditions. That’s what the Live conditions ought to be, but it’s not what’s appearing in game or on the world map.
As an atmospheric science person, I thought it would be useful to try to pin down where the variance might be between the real world weather and that depicted by MSFS. So for the last couple of days, I’ve been looking around the US for “interesting” weather that might be in the process of changing rapidly, e.g., frontal passage, etc. I compared those quantitative components that I have heard MSFS actually does take from the METARs, namely temperature, wind, and barometric pressure, along with the sky conditions that are derived from the METEOBLUE models that apparently link up to the GFS global model. I compared the weather depicted by MSFS using LIve Weather to that depicted when using REX Weather Force, which is a payware add-on that injects METAR derived conditions directly.
What I found was both surprising and not so much. First, the three basic quantitative parameters were essentially the same as that displayed by the METARS, both for Live Weather and for WF. At each airport, I set the QNH using the B key. I read both the wind and the temperature off of the PFD in the default C172. There were small differences between the METAR readings and those displayed on the G1000, but I could detect no systematic trends.
As expected, the depictions of clouds and visibility did sometimes differ from that reported on the METAR. I was surprised in one case that the WF-based depiction of 1/2 mile visibility was worse than that of Live Weather. Somehow this must have been picked up by the Meteoblue model, but WF didn’t do as well. On the other hand, a shallow marine layer in southern CA was better reproduced by WF, but almost completely missed by Live Weather.
So which is better? I think the answer to that is “It depends.” At no point did I detect the sort of gross errors that have been described by others in this post, but that doesn’t mean they were wrong in their assessments. Live Weather seems good a creating an authentic representation of some phenomena, but not others. Weather Force, on the other hand, is very good at interpreting the METAR, but in some cases it may not be as good as getting MSFS to give a good subjective depiction of the sky conditions.
MSFS has come a long way in terms of providing an authentic depiction of many weather phenomena, but there’s still room for improvement. Users can pick and choose which methodology they prefer based on personal perceptions. Choice is good.
You should try conducting these tests at different times throughout the day. The issue we’re talking about in this thread is about the live weather in sim being depicted with the wrong forecast hour.
For me, at around 2100Z-2200Z each day, the live weather in sim begins to show the weather for 0900Z the next morning, completely skipping almost 12 hours of live weather. So yes, it’s still possible to get accurate live weather, but only for a certain portion of the day between 0900Z and 2200Z.
To complicate matters, the METAR data being used is still current at these times so you essentially have a mix between current METAR data (temp, winds, pressure) and future meteoblue data (cloud, precip).
Yes, I was just able to confirm that observation, although it took me a while to find a few airports with sufficiently different weather conditions now (2300Z) and predicted the next day, representing only a very small fraction of the total available stations. Just to be clear, what I see is the correct quantitative atmospheric parameters (wind, baro, temp), but the wrong sky conditions. I also checked the METEOBLUE sky conditions and those were correct, which taken together does suggest that there are certain times every day at which MSFS servers start sending out sky conditions that are ahead of real time, so this is NOT a METEOBLUE problem.
In a sense, this is a bit of solved problem, don’t you think? Just use REX Weather Force. I checked the weather depiction at those two airports where LIve Weather was showing the wrong sky conditions, and WF depicts the right weather (not counting the lightning as usual–again we see that this is an MSFS issue, not METEOBLUE). There’s no need to reset computer clocks, etc. Alternatively use the freeware Unreal Weather which does something similar to WF.
I guess that MSFS has to have finite time resolution with regard to the sky conditions because the models themselves are only run a few times per day, so this may be inevitable to some extent. Some folks think it’s possible to run these supercomputer global weather models on a continuous basis, but that’s not so. Do point this out to MSFS via Zendesk, so they at least have a marker that reminds them to look at it.
I’m sorry but for myself and I know a lot of others on this forum, this sort of suggestion just isn’t acceptable. If you don’t own it already, why should people have to spend more money to get something to work, when it should already work for the product we have paid for?
This anger isn’t directed towards you, but as a community this attitude of ‘Just buy this’ to fix a problem has plagued all of the flight sims and now seems to be the same in MSFS as well.
Not only that, but the likes of REX and Unreal weather mod just inject whatever the METAR reports on a global scale across the full game, so that doesn’t even give a realistic representation of the sky conditions either.
I think we’d all just greatly appreciate it if a function worked as intended, instead of having to switch to work arounds, especially of a payware nature.
I have used REX Weather Force and Unreal WX. I’ve tried both. While both of those will inject METAR based weather into the sim, I and many others have mentioned they come with some big limitations. The two most significant ones being that when those mods load weather, it is rendered the same globally. You have no sense of what weather is around you. The second big limitation is the nature of METARS in general. Many stations report clear skies even when there’s cloud coverage so a lot of atmospheric conditions are completely missing when those mods are used.
I found REX to cause noticeable stuttering for me during transitions between different stations. Unreal seemed to work a little smoother but still isn’t ideal. The big advantage those mods have that MSFS doesn’t is being able to look up exactly what the weather will be in game. This is a huge limitation with the built in weather. The info available on the world map is not specific enough to be all that helpful and you can’t access it during flight.
The folks that think it’s possible are MS and Asobo who designed this system. Credit to them for getting the system working half way. Sometimes it takes a bunch of people to figure things out and maybe all the complaints about live weather really boil down to this time/forecast hour discrepancy. How many people who’ve said the weather is not accurate maybe didn’t realize the weather actually was pretty spot on, but just for the wrong forecast hour?
After installing SU4 update it hit me… I did a storm chase tonight in Amarillo, TX. this is a “stand alone” type thunderstorm tonight.
I believe those who are saying that the clouds depicted are for the next day after approximately 20:00Z are correct. Because flying out of KAMA tonight I’m seeing lightning, correct winds, and SOME clouds, but the cloud tops are like 6,000 feet and scattered, with lightning all around.
I think I’m seeing everything from current weather including lightning - EXCEPT the clouds and precipitation! We’re SO close, but not quite there yet! If Asobo could just align the clouds to the rest of the weather for the 20:00Z bug, I think we’ll have it fixed!
I agree, it’s pretty dark when overcast, especially in the cockpit view. It’s made me raise the brightness on my monitor (TV). I think the auto exposure that happens- in cockpit vs out the window is overdone.