A stack file arrives as a stack of numbers, and before any of them can be used they have to be sorted. In this subject the units do the sorting, and they do it reliably — which is the one mercy of a field that otherwise runs on fitted constants.

Out in the atmosphere: the lapse rate $\Gamma$ — capital gamma — is how fast the air cools with height, in °C per kilometre. Compare it against the dry adiabatic rate of 9.8 and you have the stability of the whole airshed in one comparison. Faster than 9.8 and the air is unstable and mixing hard; slower and it is stable; NEGATIVE — warmer above than below — and you have an inversion, the condition in which plumes flatten into ribbons and ground-level concentrations climb.

At the stack: the exit velocity $v_s$ is the speed the gas leaves, in m/s, and the subscript s means 'stack' everywhere in this chapter. The draft pressure $\Delta p$ is the pull the chimney develops on its own, in pascals. The buoyancy flux $F$ is an intermediate parameter with the unmistakable units m⁴/s³ — nothing else on any sheet wears them — and the plume rise $\Delta h$ it produces is a plain height in metres.

In the duct: a volume concentration is ppm, parts of pollutant per million parts of gas counted by volume, which is what an analyser natively reads. A mass concentration $C$ is mg/m³, which is what most permits are written in, and the bridge between them needs the pollutant's molar mass. An emission rate $E$ is g/s — a concentration multiplied by a flow — and it is the source strength every dispersion model asks for.

One nugget worth carrying. A concentration says how THICK the gas is; an emission rate says how MUCH is going by. A narrow stack and a wide one reading the same ppm are not the same source, and the difference between them is the reason dilution air was ever regulated at all.