Discrepancy in altitude reading

[pinepix62 63]

I have not been flying a lot recently and and when I flew today I noticed a discrepancy in the altitude values displayed in the sim.

I started at close to sea level from EGHC and the altimeter show the correct field altitude when I pressed the B key. I set the autopilot to 6000 ft and when I reached 6000 ft I noticed that my actual altitude was 5742 ft. I pressed B and checked the barometer setting and all was in order.

LNM reported the same values.

I repeated the test with the default Bonanza, the C310 and the Kodiak and got the same results. I pressed the B key several times and checked the barometer values to make sure it was all correct.

Has anyone else noticed this. I am not sure if the altimeter is wrong or if the actual altitude is not being reported correctly.

[Bob Scott 7,715]

It's more than likely caused by the temperature deviation.  Cold temps (below ISA) cause the actual altitude to be lower than indicated.

As an approximation, your actual altitude will vary by ~4% per 10 deg C deviation from ISA at ground level.  We know that the temp dev at 6000 ft is -8 deg C, but don't know what the ground temp dev is from what you have shown.  If we assume that the temp dev is also -8 deg on the ground, that would account for a nearly 200 ft delta between indicated and actual altitude, which is in the ball park of what you're seeing.

There are cold weather altitude correction procedures that apply when airfield temps are below 0 deg C (ref PANS-OPS Vol 1 Ch 4).

[pinepix62 63]

Thank you very much for your reply. I have been flight simming for many years and this is the first time I noticed this. Learned something new.

Just out of interest I repeated the flight from a nice and warm FACT flying the C130 and this time the altitude values were much closer together.

[jcomm 5,890]

It's a bug, either in the weather model or in the instruments logic...

If you have your AP set for 6000' and it is using the barometric altitude then you will have to read exactly 6000' in the ALT when the AP levels off the aircraft, unless the altimeter is a radio-altimeter and the surface bellow is flat (like sea) or GPS alt is being used.

I am yet to properly test that MFS correctly models geopotential height variation in ISA++ and ISA-- scenarios, but the problem in RL is that you see everything indicating "correctly" your altitude - the one you think you're actually flying at if it was an ISA day, but in reality you're flying lower or hgher...

If he altimeter in MFS's C172 shows less then 6000' when your AP levels at 6000', then either the altimeter setting is wrong, or the AP is using other than barometric pressure, or there is a bug in the variables being read by the sim or in the troposphere model.

Edited December 6, 2022 by jcomm

[martinboehme 492]

@jcomm I think you misunderstood what the OP was saying. The altimeter in the aircraft does indeed indicate 6000 feet (see screenshot in the first post), but the actual (geometric) altitude shown in the debug screen is lower. As @Bob Scott explains, this is due to a temperature deviation from ISA.

[jcomm 5,890]

40 minutes ago, martinboehme said:

@jcomm I think you misunderstood what the OP was saying. The altimeter in the aircraft does indeed indicate 6000 feet (see screenshot in the first post), but the actual (geometric) altitude shown in the debug screen is lower. As @Bob Scott explains, this is due to a temperature deviation from ISA.

Ah!  I missed that yes! My bad 😞

Thank you for calling my attention to it - then it looks OK, I didn't realize he was using the debug window and geometric alt!

Great!!! - and this is in the good way to prove that indeed MFS models geopotential height variation with T.

I still have to test the kind lapse rate for different deltaT.

Thanks!

Edited December 6, 2022 by jcomm

[martinboehme 492]

Agreed - I'm also happy to see that MSFS gets this right! 

[MattNischan 3,778]

Yes, we implemented this in SU4 and SU5, and were the first consumer sim to do so by a long time margin. You will get extremely accurate non-ISA atmosphere altimetry errors vs pure geometric altitude when in Live Weather. It's quite neat!

[Alti 63]

i learned in Flightscool: "in winter the mountains are higher" because of atmospheric lower temperature!

[JRBarrett 3,038]

2 hours ago, MattNischan said:

Yes, we implemented this in SU4 and SU5, and were the first consumer sim to do so by a long time margin. You will get extremely accurate non-ISA atmosphere altimetry errors vs pure geometric altitude when in Live Weather. It's quite neat!

I have tested this many times and it is indeed spot on. In summer especially, when the upper atmosphere is often ISA +15 (or more) geometric altitude will often be 1000 to 1500 feet higher than indicated altitude, and MSFS models this very well when Live Weather is active.

Edited December 6, 2022 by JRBarrett

[jcomm 5,890]

3 hours ago, MattNischan said:

Yes, we implemented this in SU4 and SU5, and were the first consumer sim to do so by a long time margin. You will get extremely accurate non-ISA atmosphere altimetry errors vs pure geometric altitude when in Live Weather. It's quite neat!

Welllzzzz...

Aerowinx PSX had done it long ago 🙂

Even DCS World and IL-2 Great Battles do it even if in a rather simplistic way...

I believe there's even a dynamic weather alternative for FlightGear which does it...

and.... well, XP12 does it too now 🙂

[pinepix62 63]

8 hours ago, MattNischan said:

Yes, we implemented this in SU4 and SU5

OK, makes sense, I never noticed such a big deviation as I mostly fly low and slow in the warmer areas of the world. Good to know it is not a bug with the sim or the aircraft.

[Farlis 2,277]

11 hours ago, MattNischan said:

Yes, we implemented this in SU4 and SU5, and were the first consumer sim to do so by a long time margin. You will get extremely accurate non-ISA atmosphere altimetry errors vs pure geometric altitude when in Live Weather. It's quite neat!

Awesome. I just did a flight into Bozeman last week and saw that the Charts package included a cold temperture deviation chart and was wondering if I even needed that in MSFS.

Now I know, I do.