Dynamic Weather Discussion

[MBot]

The old thread on dynamic weather seems to have been deleted yesterday for some reason, so let's get this going again with some question.

 

-Why is wind increasing from the center to the edge of a pressure system? The pressure gradient seems to be more or less the same across a system, therefore also the wind speed should remain the same. Yet in DCS, wind speed at the edge is often multiple times the wind speed in the inner area of a pressure system.

 

-What determines could generation? Neither air masses with different temperatures/densities (fronts) or humidity seem to be variables in the dynamic weather system.

 

-Why does a low pressure area have clear skies at the center and clouds towards the edge? The pressure minimum at the center would usually correlate with the most rising air masses, which is the cause for condensation and cloud generation.

 

 

I am trying hard to get a better understanding of the dynamic weather system in order to write persistent and evolving weather into my dynamic campaign engine. Yet I am having big difficulty understanding the mechanics behind it. A lot does not seem to make much sense.

[DarkFire]

While I agree totally that it would be helpful to understand the dynamic weather system in DCS world, in terms of realism unfortunately I don't think there's much point. Attached is the current (0100 hours GMT on 01/03/2016) weather plot for the Black Sea.

 

It's very difficult to read, but essentially the current weather over the Caucasus area appears to be controlled by three weather systems: one of them is over the former Yugoslavia (Serbia, Bosnia etc), the second is over Denmark and the third is over... Iran?

 

Looking at a similar surface pressure map for Nevada, the minimum distance between areas of high & low pressure appear to be about 2-3 x the DCS map diameter apart.

 

Going back to the Black Sea weather charts, what you say about wind speed across pressure differentials appears to be correct - to a first approximation they should be largely uniform.

 

While the DCS World weather system is good for representing static weather patterns I'm not sure that the dynamic system has a realistic scale built in to it.

[MBot]

Scale isn't actually a problem. When controlling the variable manually in the mission file, you can easliy define a Low 2000 km to the west and a High 2000 km to the east. With such distances, the pressure gradients would be reasonable even with large pressure differential. Unfortunately you still get excessive wind speeds towards a system's edge and clear sky even in a 960 hPa Low.

 

 

Also, the system seems to be pretty unstable. I get a DCS crash in about 1 in 10 attempts to load a mission with dynamic weather.

[AG-51_Razor]

I wonder how difficult it would be for ED to allow the mission builder to select different areas within the map to assign different wind directions/velocities, cloud levels and visibilities? I would much rather have a system where I could select/control the weather in my mission than to leave it to chance. And what I mean by that is, I would like to be able to choose a particular weather condition in one section of the map and a completely different set of conditions in another section of the map. I don't know if that would be easier than the "dynamic" system they have in place right now, or even if it is achievable but it would certainly be preferable IMHO.

[DarkFire]

Scale isn't actually a problem. When controlling the variable manually in the mission file, you can easliy define a Low 2000 km to the west and a High 2000 km to the east. With such distances, the pressure gradients would be reasonable even with large pressure differential. Unfortunately you still get excessive wind speeds towards a system's edge and clear sky even in a 960 hPa Low.

 

 

Also, the system seems to be pretty unstable. I get a DCS crash in about 1 in 10 attempts to load a mission with dynamic weather.

 

Ah, that's very useful to know! I always assumed that the definition of the pressure system was limited by the size of the in-game map. Interesting... I might have to experiment with this...

[Northstar98]

I wonder how difficult it would be for ED to allow the mission builder to select different areas within the map to assign different wind directions/velocities, cloud levels and visibilities?

 

Another thing you could on top is set time intervals for each location as well say at mission start the weather is a bit cloudy, no precipitation and a bit of wind, then say select a different time (either entering the time in or selecting say xx hours, xx minutes etc) the weather changes to this. Or do what Arma has which is starting weather, ending weather and the time period between (though for longer sorties and missions I think there needs to be more progression, which may mean the user has to enter multiple weather settings for more frequent time intervals

Edited March 1, 2016 by Northstar98

[Grimes]

I was able to salvage some of the data from the old post.

 

Here is a quote from EvilBivol-1 from all the way back in 2011 from that thread. I honestly don't know if the information is still accurate. From the looks of things the weather generation stuff hasn't changed, but I don't know if how the values are interpreted could have changed since that data is likely hidden in a dll.

 

As it is designed now, the basic principle to follow is that Low pressure systems tend to produce clouds, while High pressure systems tend to produce clear skies. Low pressure systems are cyclonic, where the center point has the lowest pressure and winds (and clouds) tend to form toward it. The opposite is true for High pressure systems. The pressure excess value determines the variation in pressure between the center and outer edge of the pressure system. The greater the variation, the stronger winds it will cause around the system.

 

The trickiest part in using the dynamic weather system is to try to generate cloudy conditions without excessive winds. I've had good success by placing two Low pressure systems close to each other and making one "dominant" (greater pressure excess value) over the other. This can generate a lot of cloud cover, but minimize wind force, because there is no High pressure system to suck the air from.

 

I don't know a whole lot of the science and expected results behind the weather, but I'll try and get other useful data from that thread into a wiki article or something.

 

Also, the system seems to be pretty unstable. I get a DCS crash in about 1 in 10 attempts to load a mission with dynamic weather.

 

I'll be on the look out for that. Did you experience it in both 1.5 and 2.0?

[MBot]

Thank you Grimes. I haven't actually used 2.0 a lot, but i *think* I also saw crashes there. Certainly in 1.5.

 

EB's comment that clouds form towards the center of low pressure systems sounds correct (if we disregard that most weather effects by pressure systems are actually caused through fronts, which are not simulated in DCS). In DCS it seems to be the opposite though with clouds forming towards the edge. His comment about winds is confusing me even further, as I do not see the rational of winds being low in the center and strong at the edge.

 

Real meteorologists please feel free to correct me, this is the basics as I remember them from meteorology classes.

[dooom]

no crashing on DoW with dynamic weather - we are using it in roughl 3 or 4 of our missions.

 

i really really really do wish we had a library of dyn weather templates for download though. it is unfortunately difficult to tease out the nuances of setting up a decent pattern for a mission - well labelled, available templates spanning different map areas woudl be most welcome indeed.

[Shahdoh]

Have been using dynamic weather on the VA server for quite some time. No issues while it is up and running. But almost all clients have a crash to desktop when trying to reenter the mission either from another server or leave mission and rejoin. Advice we have given for the moment is to just exit DCS completly before coming back in and everything is fine after that.

[towsim]

Going along a physical model is a huge challenge. In ATC tower simulations we made the experience, that a physical model has two disadvantages. The first is, that the number of variables, interacting each other, makes it almost impossible to predict a result. It requires a deep understanding in meteorological physics to create a specific weather situation at a given time. The second is, that in a serious simulation, a situation must be reproducible. The latter is difficult, if too much variables have to be considered. A standard situation in a military environment is a takeoff in pretty VMC, executing a mission, while ceiling and visibility decreases on return. The pilot has to recover in IMC and to change over to an IFR flight plan. We found a solution to create weather management files, which can be bound to any exercise (mission). It gives the possibility to use a specific mission in different weather situations. Even this has a disadvantage. Such a system can be configured with completely illogical parameters, like defining heavy snow precipitation with an air temperature of +35°C. Even here, a basic knowledge of meteorology is needed. The current version of static weather in DCS gives the possibility to define some parameters to create a situation. It would be a big advantage, if the parameter could change over the time and this in different geographical areas. The attached image shows a weather management editor which is used with a professional ATC tower simulator (TOWSIM). It lets the user change the variables by moving the points in the diagrams directly. The result can be stored in a weather management file and later bound to any mission. I do not expect ED to change their simulation model, but it can demonstrate a rough direction for further developments.

[Grimes]

Oh yeah. I started a wiki article. I tried making a gif of different settings, suffice to say it didn't go well.

[Stonehouse]

Found this lurking on my HDD, not my work thank DAWS I am just the messenger. Will do some more looking around to see if I took any notes or other downloads from the old thread.

 

<edit> and one from yakbeard

 

Cheers,

Stonehouse

DAWS Weather-pkg-rev.03.zip

DCS Dynamic Weather v1.pdf

Edited April 22, 2016 by Stonehouse
found a later version

[chromium]

DAWS is an acronym and it has nothing to do (now) with dynamic weather. The mod is made by AMVI and current version is here:

http://forums.eagle.ru/showthread.php?t=144855&highlight=DAWS

 

To be honest I found a way that will update current static weather based on icao code at every multiplayer server creation, regardless of mission editor settings... but it's in a very early stage.. I will release an update for DAWS Weather when it is ready.

 

To keep the discussion, I'm looking in zyGrib software that show something similar to DCS Dynamic system and asked on their forum (http://zygrib.org/forum/viewtopic.php?f=4&t=954&sid=a82bacef9d095ae875d72c08d617a0a7) if there is a way to extract cyclones information from the downloaded grid. If so... you may imagine the result. Sadly i found it currently not possible :(

 

The yakbeard document is fantastic!

Edited April 22, 2016 by chromium

[Vedexent]

You need not move to a physical modelling system.

 

There are 3rd party add ons (for another flight sim made by a company that probably wrote your operating system ...) which imports historical (or even real time) real-world meteorological data.

 

Don't bother trying to simulate weather; just replay it.

 

I'm sure there are other technical issues - and perhaps licensing issues for meteorological record access - with this approach, however.

[chromium]

Don't bother trying to simulate weather; just replay it.

 

That is the point. zygrib do not simulate anything, is a database. But sadly I'm not a programmer and neither skilled or sufficient time do dig in so deeply.

[Stonehouse]

Thanks Chromium, apologies for the mix up in accreditation. I just couldn't remember where I have got that document from.

 

DAWS is an acronym and it has nothing to do (now) with dynamic weather. The mod is made by AMVI and current version is here:

http://forums.eagle.ru/showthread.php?t=144855&highlight=DAWS

[chromium]

No problem. I hope that in a certain point we will be able to find cyclones data somewhere. If so, I will try everything to convert DAWS Weather into a dynamic weather mod.

[Stonehouse]

Saw this today too. DCS 1.x so may not be fully relevant but background info etc may be useful http://forums.eagle.ru/showthread.php?p=1133542#post1133542

[chromium]

Ok let's see if something is possible to do. Let's say that with a certain degree of error, we might achieve a working dynamic weather mod IF cyclones informations can be retrieved somewhere.

 

What does we (realistically) need?

 

those ingoformation:

-range of influence (size...)

-center point coordinates

-pressure at center point

-ellipticity

-rotation speed

for every cyclones/anticyclones in a range of... 2000 km? from map coordinates 0,0.

 

zygrib software can show those data. it use grib files. Can someone find a method to:

1. gather those data from NOAA site

2. read & extract data from the grib file?

 

We may, for start, assume that the range is always 500 km or something similar. the wind speed = 0 information should give us the coordinates of each cyclone. reading wind direcition 10km side east and side west of each center will give us rotation speed with some approximation. Ellipticity... dunno. Let's say = 1 for testing purposes.

 

Ideas?

[wilbur81]

I know I'm surely ignorant, but I don't see why the Dynamic Weather feature can't just work like trigger zones? Choose the size, type, etc. of system, put it on the map WHERE YOU WANT IT, and viola.

[DarkCrow]

I would love to have more control over the weather around the map. I've been trying to use the dynamic weather for the Huey campaign I'm making because of hearing stories where clouds roll in in mountainous terrain and the helicopter pilot has to wait for it to clear or find another way to pass, etc. I was able to do this with the dynamic weather but for some reason the clouds disappear when they get near the edge of my monitor killing the immersion. Or sometimes they just pop up right in front of me then disappear again. I believe this is a huge part of aviation and really important in mission design.

 

 

Sent from my iPhone using Tapatalk

[showtime817th]

I lost my hard drive a few week ago and I've been searching this forum for couple weather temples and I beleave MadTommy was the author for Batumi front and NE-SW front. I would appreciate if someone as the old link or the files to share.

 

Thanks.