Pressure units are simple; the reference point is not
Most pressure conversions are ordinary multiplication — psi to kPa, bar to atm — and cause no more trouble than any other unit pair. The genuine, safety-relevant mistake in pressure measurement is not a wrong conversion factor at all. It is using two numbers that measure pressure relative to different baselines and treating them as comparable.
Gauge pressure and absolute pressure are not the same quantity
Gauge pressure is measured relative to the surrounding atmospheric pressure — a gauge reading of 0 means "the same as the air around it," not "no pressure." Absolute pressure is measured relative to a true vacuum. The two differ by whatever the local atmospheric pressure happens to be, roughly 101.3 kPa (14.7 psi) at sea level, and that offset is exactly the size of the mistake if the wrong one gets used.
Worked example: the tyre
A tyre gauge reading "32 psi" is reporting gauge pressure: 32 psi above whatever the atmosphere outside the tyre is doing. The tyre's absolute internal pressure is therefore about 32 + 14.7 ≈ 46.7 psi. Tyre manufacturers, tyre gauges and the pressure sticker inside a car door all specify gauge pressure by convention, so this rarely causes a real error on its own — the mistake shows up when someone takes a gauge figure into a physics or engineering calculation, the ideal gas law, for instance, that requires absolute pressure, and forgets to add the 14.7 psi back in.
Worked example: the barometer
Weather reports run the opposite convention. "1013 hPa" on a forecast is an absolute pressure figure, more precisely a sea-level-corrected absolute pressure (see below), not a pressure relative to some reference. Standard atmospheric pressure, 101,325 Pa exactly, was fixed as "1 atmosphere" in 1954 by the 10th General Conference on Weights and Measures, and it is the same 101.325 kPa that has to be added back onto a gauge reading to obtain an absolute one.
The unit landscape
| Unit | Exact relationship | Typical use |
|---|---|---|
| Pascal (Pa) | SI base: 1 N/m² | Scientific/engineering base unit; rarely quoted bare, being so small |
| Bar | 100,000 Pa exactly | European tyre pressures, weather maps, diving |
| Atmosphere (atm) | 101,325 Pa exactly | Chemistry, diving, defining "standard conditions" |
| PSI | ≈6,894.757 Pa (from lbf/in²) | US/UK tyre pressure, hydraulics, compressed gas |
| mmHg (conventional) | 133.322387415 Pa | Blood pressure, older barometric readings |
| Torr | 101,325/760 = 133.3223684… Pa | Vacuum systems, physics |
mmHg and torr are close enough, differing by about 0.14 parts per million (mmHg = 133.322387415 Pa vs torr = 101,325/760 = 133.322368421… Pa, a difference of 1.899×10⁻⁵ Pa on a base of about 133.3 Pa), that most references treat them as interchangeable, but they are not defined identically: mmHg derives from a specific conventional mercury density combined with standard gravity, while torr is defined as an exact 1/760 fraction of the standard atmosphere. Vacuum physics, where the distinction occasionally matters, tends to use torr specifically because its definition does not depend on the physical properties of mercury at all.
The altimeter-setting problem
Aviation depends on absolute pressure for altitude, and it depends on that reference being current. An aircraft altimeter is, mechanically, a barometer calibrated to read altitude rather than pressure — it needs to be told the current local sea-level-equivalent pressure, the altimeter setting or QNH, to convert its raw pressure reading into a correct altitude above mean sea level. Fly from a region of high pressure into a region of lower pressure without updating that setting, and the altimeter keeps reading against the old, higher reference, so it displays an altitude higher than the aircraft's true altitude. The standard pilot mnemonic is "high to low, look out below": the error runs in the dangerous direction, indicating more clearance from terrain than actually exists. As a rough working figure, a 1 hPa error in the altimeter setting corresponds to about 30 feet (9 m) of altitude error near sea level; the exact figure varies with altitude and temperature, because atmospheric pressure falls off logarithmically rather than linearly with height, but 30 ft per hPa is the standard rule of thumb pilots are trained to use for a quick mental check.
Why weather pressure is corrected to sea level at all
Raw, or "station," atmospheric pressure falls with altitude — in the US Standard Atmosphere, about 83,430 Pa at one mile (1,609 m) of elevation versus 101,325 Pa at sea level, roughly 18% lower, with no change in weather at all. (Denver's real station pressure, around 835–843 hPa, is consistent with that figure.) Reporting raw station pressure would make every high-elevation city look like it was permanently sitting in a severe low-pressure system, so meteorological reports correct station pressure to what it would read if the station were at sea level, using the station's known elevation. That sea-level-adjusted figure, not the raw sensor reading, is what "1013 hPa" on a forecast actually represents: a second, separate correction from the gauge/absolute distinction above, and one more reason a raw pressure sensor at a given location and a published weather figure for that same location rarely match exactly.
Negative gauge pressure has its own vocabulary
Gauge pressure can go negative, when the measured pressure is below the surrounding atmosphere, and trades that deal with this routinely give it a separate name, "vacuum," rather than writing a negative sign. Refrigerant technicians evacuating an air-conditioning line before recharging it watch a gauge reading in inches of mercury vacuum, working toward roughly 28-30 inHg vacuum (close to, but short of, a perfect vacuum) before deeming the line dry enough to charge. That figure is still a gauge measurement, just counted downward from atmospheric instead of upward from it, and it converts to absolute pressure the same way the positive case does: subtract it from local atmospheric pressure rather than add it.
A rule that resolves most everyday confusion
If a pressure reading comes from something designed to sit inside a larger ambient pressure, a tyre, a scuba tank, a hydraulic line, assume gauge. If it comes from something describing the atmosphere itself, a weather report, a diving depth table, an aircraft altimeter, assume absolute, or at minimum sea-level-corrected absolute. The two conventions are both entrenched and both reasonable within their own domain; the error only happens when a figure crosses from one domain into a calculation built for the other.