For example:
An engine tested on a Dyno makes 400BHP at 15°C at 1013Mb (millibars) air pressure, at sea level.
The same engine is now tested at 0°C, at 1030Mb at sea level. You would expect the engine to make more power, say 450BHP due to colder, denser air.
The engines efficiency has not changed, no internal or external components have been changed, but due to the simple fact that an engine is an air pump, by having more air (colder, higher pressure air, hence more dense at the intake) means it moves more air per cycle than in the first test at 15°C at 1013Mb.
Take the same engine up a big hill, to say 5000ft above sea level, on a really hot day....
The conditions are now +30°C, with an air pressure of 990Mb...
It now "only" makes 350BHP, as the air at the intake is really hot, and really thin! So again, no internal or external components have been changed, but per cycle, there is far less air to move due to the heat/density than at the first test, and far, far less than the second test.
In all three examples, the engine is the same, but gives very different figures on the same dyno purely down to external atmospheric condions...
The point of the Dyno air temp & pressure sensors are to allow the computer to apply a "correction factor" to the conditions of the day, so theoretically, if the said 400Bhp engine is tested in any of the above scenarios, or in any conditions anywhere in the world, it will read 400Bhp on the dyno, as the computer is comparing the actual conditions on the day to the "ideal" conditons that it uses as a base for all calculations, as in the first example.
