Inconsistent temperature

[vonotha]

The displayed temperature seems to be in disagreement both with Weather Editor and weather parameters returned by SE_GetWeather.

Other weather parameters (sea state, rain, undercloud) seems fine.

No Custom Environment Zone is defined at this moment (it was before but was deleted)

weather.jpg (59.15 KiB) Viewed 903 times

[Parel803]

Goodmorning,

Believe the temperture is depending on the Latitude you are at. Temp goes up or down with North or South.
If I misuderstood the question I do apologize.

best regards GJ

[vonotha]

You are absolutelly correct, the temperature (but no other weather parameters) change depending on the measurement place.

But what is going here? Two measurements taken close to each other - one in France and one in Germany. 10 C difference.

And I was not aware of the time difference between those two countries?

w2.jpg (25.21 KiB) Viewed 875 times

w1.jpg (14.1 KiB) Viewed 875 times

[Parel803]

You are, sorry cannot help any further.

Just to be sure I'll check nect time I'm iver there

And I was not aware of the time difference between those two countries?

[boogabooga]

It's not by latitude. It's by the time of day (Day, Dawn, Dusk, Night).

Night is the coldest.
Dawn or Dusk -> night temp. +10
Day-> night temp. +20

The "Average Temperature" slider sets the Dawn/Dusk temperature.

[Parel803]

Thanks, did not think of that changes.
regards GJ

[Deserere]

should we consider this as "working as intended" then?

[boogabooga]

I don't know, does a 20 deg. C day-night variation seem right to you, for most of the world?

[thewood1]

I live in the Boston area. Its very plausible. I'm not sure it matters that much either way.

[KnightHawk75]

I don't know, does a 20 deg. C day-night variation seem right to you, for most of the world?

Good question. hmm, I can definitely think of more than a few places where it wouldn't be accurate at 20c but would be at 10c, others 15c. Examples of the first that come to mind are hawaii (~10c), DC(~10), Hong Kong (<~10) and Brazil in total (~10 ), Moscow(~10). Most of the US east cost in say the month of july rarely exceeds ~10, midwest ~15, and west ~20(and even then it's not everywhere). Now as for 'most of the world', i simply don't know.

https://www.climatestotravel.com/climate/brazil
https://www.climatestotravel.com/climate/hawaii
https://www.climatestotravel.com/climate/hong-kong
https://www.climatestotravel.com/climat ... sia/moscow

Boston for example would be covered by ~10C as well per ( this , and that ) fwiw.

Maybe a better baseline would be 10C? With adjustments (keeping it fairly simple) based more on land-cover types, lat and elevation which are common factors for which the game has some information, and I don't have some world wide implementation of it to match with thousands of cities to check if matches known ranges, but generically how about the the following:

Possible adjustments to baseline offset figure of 10C differential for "night".
Say +5C for:
Cropland natural vegetated mosaic
Barren or Sparsely Vegetated.
Grasslands
Snow and Ice?

Say -5C to for:
Water?

Also +5C for:
elevation > 3000ft? (I'm just picking semi arbitrary number)
Also +5C for:
lat > 45N | lat < 45S (I'm just picking arbitrary "higher\lower" lats per some descriptions about general trends while searching around)

So you could get a value as high as 25 or as low as 5.

Again I didn't run the proposal against data worldwide, cause I don't have it handy. I but I do think it would represent something at least a little closer to irl and might be better than a hard 20C with out any variance. Is it worth it to run that extra code on every read of lat\lon point? I don't know, but if not, maybe a flat 15C for the world might be more in line with over-all average from just spot checking a dozen places around the globe than 20, which isn't enough data points I grant you. Anyway food for thought. /2cents

[boogabooga]

I live in the Boston area. Its very plausible. I'm not sure it matters that much either way.

We are talking about 20 degrees Celsius. That's 36 degrees Fahrenheit. How often does an 80 degree Boston summer day get to 44 at night? I'm sure it happens (today is an example, actually), but probably accompanied by a cold front moving in. Not every day.


I think it matters in the sonar model and IR(?).

[thewood1]

We had a recent day just this year that went from about 50 to 0 at night.

[boogabooga]

The weather model in my A2A BVR quick missions basically undoes the day-night variation, so it's something that I have noticed, but I certainly expect the developers to have bigger fish to fry.

[Dimitris]

Logged #15039