Lightning in Live Weather - Excessive and where Lightning should not Occur

Perrry · October 14, 2021, 10:09am

The problem with thunderstorms is that it’s hard to predict where they will occure and they can only last for a couple of minutes and they can suddenly appear. I think they need to have like a trigger in the sim like if the meteoblue predicts or METAR reports thunderstorms and inject it in the sim the simulator needs decide if it should simulate a thunderstorm in dynamic and varied way. This should be randomly because as i said they are hard to predict. Sometimes we get a predicted thunderstorm sometimes we don’t as it is IRL. Everybody that uses the live weather in MP needs to see the thunderstorm at the same time though. They should not inject the meteoblue reported clouds when the thunderstorm is happening. The sim needs to create a thunderstorm by itself and the injected weather that was before the thunderstorm should transission smoothly to the thunderstorm. That could even make the turbulence around thunderstorms more realistic with the pressure impacting the air-movement. That would make it more realistic to see the build up of the thunderstorm aswell. The thunderstorms has different stages and that could make us avoid them too if we want to.

Edit: If METAR reports thunderstorms it should always simulate a thunderstorm because it actually has been a thunderstorm there recently or maybe it could always just use the METARS that reports thunderstorms to tell the sim to simulate them. Meteoblue is only prediction of thunderstorms thats why we get them everywhere they are predicted and lasts as long as there is a prediction of it. they will not occure on all those predicted places IRL.

I would be happy if they could simulate buildup of a thunderstorm in a realistic way. Maybe that’s to much to ask for but i think those things they need to focus on to give us an immersive experience.

hobanagerik · October 14, 2021, 11:05am

I don’t think the wind gradient is right either. They halve the wind speed at ground level, likely leading to the stories of lurching take off rolls, and landings.

I noticed this thread just had the “need-your-help” tag added to it, as I have that set to alert me.

Is there anything specific we need to do? Ah, I’ll wait…

Sleevezipper · October 14, 2021, 12:25pm

Someone is going to tell how we can help fix this

OlieTsubasa443 · October 14, 2021, 12:26pm

Hello everyone,

I just added the “Need your help tag” to help us repro the issue. For any weather related problem, please provide:

Accurate time of day (in local or Zulu time)

Date

Location

Coordinates

Were you on live weather?

Current real-world METARs for any nearby airports

Screenshot if you have one

Screenshot of the Meteoblue current conditions for that area

If any other elements are needed, I will let you know.
Meanwhile I’ll gather latest information given here.

Thank you for your help.

CONeal81 · October 14, 2021, 12:33pm

6:30 PM Central Time
October 13th 2021
Chicago Ohare International
41.9803° N, 87.9090° W
Live Weather

Lightning all over when no thunderstorms were even close to Chicago.

HalberQuacky · October 14, 2021, 12:43pm

I’m 99 percent certain that at present, Live Weather is simply using the CAPE parameter in the MeteoBlue model to create lightning. CAPE is an indication of instability, and while it can indicate the potential for thunderstorms, a high CAPE value does not necessarily mean that there are thunderstorms.

The MeteoBlue smartphone app has a map option to display predicted CAPE from the NEMS global model, and I have found that if CAPE is shown at all (even less than 1000 j/kg for a specific area), then there will likely be lightning in Live Weather when you spawn an aircraft in a location where any amount of CAPE exists.

If LiveWeather is simply going to use CAPE as a predictor of lightning, they should also look at the model’s cloud prediction, and only allow lightning to occur when clouds with extensive vertical development (i.e. cumulonimbus) are also predicted for the same location.

METAR thunderstorm reports would be one way to put thunderstorms into the sim, but it really only indicates when a thunderstorm is active at the airport itself. You might have a big cell 15 miles north of a given airport, but unless it moves over the airport, it will not appear in a METAR.

Real-time NEXRAD weather radar would be the perfect way to put thunderstorms into the sim where they are actually occurring - but NEXRAD is not available everywhere. At present, the USA, southern Canada, northern Mexico, the UK and Western Europe, Japan and Australia have NEXRAD coverage, but other parts of the world do not. There is a cost factor involved as well. Real time NEXRAD radar data can be obtained from NWS servers in the US at no cost, but that is probably not the case for other countries that have NEXRAD coverage.

Obtaining and injecting real-time lightning data from a detection network like Blitzortung.org would be another possibility. That network is a crowd-sourced system, and there may be restrictions on commercial use of the data. That would not be the case with a commercial lightning detection network such as Unisys NLDN, but that data is extremely expensive, and has a limited coverage area.

SubtotalGuide81 · October 14, 2021, 12:48pm

The devs still can’t repro the issue? They are asking which kind of weather when the bug clearly states Lightning in Live Weather. If someone reports a location and the devs check a day or more later they might very well miss it.

N316TS · October 14, 2021, 12:56pm

The development team may have access to some historical weather data.
This same request was made the first time around, and I suspect that here may be further elements in the Meteoblue models that they need to analyze to see why this is happening.

hobanagerik · October 14, 2021, 12:58pm

I had it in PNG last night, so I will fly from there again tonight to see if I can experience the same weather.

HalberQuacky · October 14, 2021, 12:59pm

Here is the current CAPE map from the MeteoBlue app for the eastern US

The CAPE for Detroit KDTW, shows about 500 j/kg, and if I spawn there right now, there is indeed lightning. But, in reality, the only place on this map where thunderstorms with lightning are actually occurring is in southern Texas, where the CAPE value is shown as yellow or orange on the map. This is confirmed by current NEXRAD, and the Blitzortung lightning network.

SubtotalGuide81 · October 14, 2021, 12:59pm

Well then that could explain why they still can’t repro it, the historic data might be processed differently.

Perrry · October 14, 2021, 1:51pm

Here you have the CAPE-map on meteoblue-webpage if that is for some help.

DaveS815857 · October 14, 2021, 2:00pm

@N316TS please have them look at this. I just tested KCVG airport at 14Oct most (recent metar was from 1252z) which is in the purple area with a low j/kg value yet I had lightning as soon as live weather was turned on. Here is what it looks like for weather though: Lightning was shooting out of the blue sky above the clouds in this picture

Conversely if I spawn to an airport in an area outside any cape value in this case KBNA airport (again metar was from 14Oct 1252z) there is no lightning even with more cloud coverage:

I am convinced that @HalberQuacky is correct with CAPE. They need to investigate this and even disable lightning until a better solution can be implemented

Perrry · October 14, 2021, 2:03pm

Meteoblue should invest in an own lightning radar service And then the lightning that occured could be injected in realtime

HalberQuacky · October 14, 2021, 2:22pm

If they are going to use CAPE alone to “predict” lightning, they should (at the very least) set the threshold much higher - say 1500 J/kg or more. Right now it appears that any amount of CAPE will result in lightning, which is completely incorrect.

In the real world, even very high CAPE values do not always guarantee thunderstorms will be present. In summer, it is possible to have surface-based CAPE of as much as 2500 J/kg (but no storms) if there is a strong capping inversion at mid levels of the atmosphere.

sydicek · October 14, 2021, 2:43pm

Yeah but then there are situations where 200 J/kg of CAPE spawns a storm. Especially during the spring/autumn/winter months in central Europe in western zonal flow weather regimes - deep lows, strong cold fronts.
I’d like to see a thundersnow in the sim, but unfortunately I’m afraid we won’t be able to see that… I don’t see how to approach that particular scenario in order to simulate it well.

HalberQuacky · October 14, 2021, 3:03pm

True - we occasionally get thunder snow here in the northeastern US in strong winter storms. But, in general, generating lightning in the sim just because “some” relatively small amount of CAPE exists in the model is the wrong approach.

If I had control over the Live Weather development, I would (at the very least) set the CAPE threshold for lightning much higher than it currently appears to be, and give users a menu option to disable all lightning generation entirely if they wish - which would probably be relatively easy to implement.

Predicting the exact location and timing where storms will occur is a difficult task even for the very best weather model. A METAR “TR” report can at least indicate that an actual thunderstorm is in progress at a given airport at a given time, but that is only valid for the specific airport.

Perrry · October 14, 2021, 3:30pm

I think the most important thing is to have it realistic in the sim. It’s not that realistic to see lightnings when there is no clouds. They could start with if there is cumulonimbus clouds that reach high enough to reach freezing temp and then only strike lightnings from those clouds that i think is a good start even if they are not located exactly where they actually happens IRL. Would be good to have them located near the place they occure IRL though if possible.

sehnsedahamses · October 14, 2021, 3:44pm

I want to note that lightning should be rendered already from long distances. Real lightning can be easily seen for a hundred miles or more in the dark.

Dont know if this was discussed here before, if so, please see this post as an reminder.

Perrry · October 14, 2021, 3:46pm

Exactly what i mean, they need to be formed in the sim in some way to be realistic. They can’t just add clouds and lightning at random places in the sky as it is right now. It can be clouds in the sky then the lightning strikes where the sky is blue and no clouds. The sim needs to know where they should strike and from what that will also make sure they not strike where they not should strike i think.