The Standard Atmosphere dates back to 1962 when Earth was colder. The 1976 update didn't make changes within 50 km altitude (the tropo- and stratosphere). It assumes 15°C at sea level, decreases until minus 56.5° at 11 km (36,000 ft) and keeps that value until 20 km (66,000 ft) from where temperature increases.

Now I wonder whether 15°C at sea level is still the appropriate temperature to be set. It's a value that would be reached at 45th latitude from late fall to early spring but is it really appropriate to use a model that cold, especially since Earth has become warmer? Take the surface pressure on Mount Everest's summit for instance: in spite of an absolute elevation of 29,032 ft, even in January the summit of Everest is barometrically lower than that, it doesn't ever reach an altitude as high as FL290, i.e. a pressure as low as 4.575 psi. Mt Everest's average surface pressure is slightly below 5 psi, making Mt Everest actually slightly higher than 27,000 ft in the International Standard Atmosphere. If your plane's altimeter shows 28,000 ft, you're able to fly over the summit of Mt Everest.

Now let's take a look at the Chimborazo which is on the very equator and so the temperature doesn't change but remains the same throughout the year. Despite having an absolute elevation of 20,548 ft the barometric altitude of Chimborazo's summit is actually around 19,000 ft, again more than a thousand feet lower than its absolute elevation. That's because Ecuador's sea level temperature is higher than 15°C.

Now you may say "But around 45th latitude the average temperature is still around 15°, and closer to the poles it's even colder". However, noone lives on or too close to the poles, and main climbing and vacation seasons are in summer when the atmosphere is more stretched and mountains have lower barometric elevations in the ISA than they actually have above sea level, and so the altitude shown in altimeters isn't the same as the absolute altitude. That's why from a practical point of view, I see more sense in setting sea level temperature to 20° or 25° as more appropriate perhaps, and setting the pressure altitude from that warmer value.

Another part of the ISA which I don't quite agree with is the alleged isothermal layer from 11 to 20 km because I find it oversimplified: it doesn't exist like this. Temperature just decreases much more slowly beyond 11 km until an altitude around 16 km (52,500 ft), from there it increases again, and above 20 km just more quickly. And considering the Earth having gotten warmer, this may change upwards as well.

Did the Earth get warm enough so that the ISA should update their pressure values for an Earth with a sea level temperature at 20-25°C ? Could it be realized easily with the altimeters of aircraft to be readapted?

  • 6
    $\begingroup$ Your title question kind of belongs on EarthScience.SE. A more appropriate question for Aviation.SE would be how bigger and bigger deviations from ISA would negatively affect aviation. $\endgroup$
    – Bianfable
    Jul 4, 2022 at 8:33
  • $\begingroup$ @Bianfable Well, other than not showing the absolute altitude (if you needed it for some reason) I don't see any negative effects on aviation. The altitude shown is actually too low. It would be a problem if it was shown too high, e.g. FL300 despite Mt Everest being 29,032 ft might result in a crash into the mountain. However, the contrary is the case. If users think it's more appropriate for Earth Science, feel free to migrate the question there. $\endgroup$ Jul 4, 2022 at 8:36
  • $\begingroup$ In any case, the satellite temperature data shows the warming phase that started in the 70s stopped about 15 years ago and we may be entering a cooling phase like the one between 1940 and the mid 70s, or worse, a return to the mini ice age that bottomed out in the early 1800s. $\endgroup$
    – John K
    Jul 4, 2022 at 12:31
  • $\begingroup$ @JohnK That's not really true for surface air temperature. The Temperature Anomaly w.r.t. 1951-80 reached its highest value (1.01°C) in 2016 and again in 2020. 15 years ago (2007) it was only 0.66°C. Source: NASA $\endgroup$
    – Bianfable
    Jul 4, 2022 at 12:57
  • $\begingroup$ @JohnK But so far the cooling phase didn't start (or didn't properly as yet) and in the mediterranean there are temperatures over 35°C. That means a plane's altimeter (when set to 29.92) will show a much lower altitude than the absolute altitude it is at. $\endgroup$ Jul 4, 2022 at 14:30

1 Answer 1


Yes, it is up to date in the de jure sense that no updates were formally accepted. There are separate models for 'hot' and 'cold' days (and other conditions), but they have limited use.

There are several points to understand regarding the ISA model.

  • First and foremost, this is a deliberatly simplistic model, even for the knowledge of the day it was proposed. Its purpose is not to describe the atmosphere for scientists, but to provide a common shared standard, primarily for engineers.

    Having a common standard makes things compatible and easily comparable around the world. For example, when an aircraft is tested, the conditions on the day never match ISA. But the results are then recalculated to ISA, so that the results of other aircraft, and even of the same aircraft tomorrow, were directly comparable. And given that the model is simple, it is relatively easy to do.

  • The surface temperature of 15°C was chosen as the closest 'round' number to the global average surface temperature on Earth. In fact, at the time the standard was developed, this temperature was closer to 14° C, and now it probably matches better.

  • In addition, the model is deliberately middle-latitude one. Typically, at the equator, despite higher surface temperature, temperature falls faster (and for longer), making the stratosphere typically colder. The opposite happens at the poles. The ISA model tries to average both cases out.

  • The model is noticeably conservative with regards to pressure (as you pointed out yourself), providing an additional safety margin for aviation to clear obstacles in most practical conditions. (I'm not sure this was a specific reasoning for choosing the parameters, though).

  • As always, changing an established standard, even a deficient one, is a very big deal that can cause numerous problems and even accidents. Standards are always conservative.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .