National Physical Laboratory

How do I adjust my barometer? (FAQ - Pressure)

The answer to this question depends on the type of barometer in question, particularly whether its pressure sensing mechanism is influenced by gravity or not, and what you are trying to get it to measure - the air pressure immediately surrounding the barometer or the pressure at another height, such as mean sea level.

Altitude effects

All barometers, whether they are mercury-filled, aneroid or other designs, are affected by altitude because atmospheric pressure reduces with vertical distance from the Earth. If a barometer is moved upwards from nominal sea level its readings should reduce roughly by 1 hPa (1 mbar) for every 10 metres increase in altitude but note that the rate-of-change of pressure with altitude itself reduces significantly and non-linearly with altitude.

Mercury barometers are doubly affected by altitude because their readings (ie the length of their mercury columns) are also dependent on the local value of gravitational acceleration - which itself reduces with altitude - to which they are subjected. A gravity correction is undertaken separately from and in addition to altitude-induced direct pressure effects and is described below.

All barometers measure the pressure at the level of their internal sensor - the surface of the cistern mercury in the case of a mercury barometer. They do not measure the pressure at sea level, as is sometimes implied, unless they happen to be at sea level. For weather forecasting purposes, however, and to enable pressure values at different locations to be compared, pressure values issued by weather forecasters are usually mathematically corrected to an arbitrarily agreed datum height - mean sea level. In other words, the pressure values published or broadcast are not those directly measured by barometers - which are rarely sited at sea level - but rather the pressure values that those barometers would have measured had they been sited at (ie mostly lowered to) sea level.

So if your barometer is not sited at sea level and you wish to compare its readings with weather forecast values you will first need to calculate and apply an altitude correction. This is in addition to the gravity-related height issue applicable to mercury barometers mentioned above and it is a bigger and more significant correction too.

With rare exception, mercury barometers cannot be adjusted for altitude - their scales are fixed and the correction has to be applied by calculation. (The exception was a 1940s instrument which incorporated a mechanical computer with a latitude adjustment that automatically calculated the corresponding value of g and varied the barometer's pressure readings accordingly. Even this instrument, however, did not compute pressure values corrected to sea level.)

Some domestic aneroid barometers have a scale or pointer whose position and hence reading can be adjusted. This is a fairly crude method, adequate for relatively inaccurate domestic purposes, that allows an instrument's readings to be 'offset' to bring them in line with those of a pressure value from a trusted source. If that value is itself corrected to sea level then the subsequent readings of the barometer similarly will be those that it would have indicated had it been at sea level. It should be stressed, though, that this adjustment procedure cannot ensure that the barometer will change its readings in line with changes in atmospheric pressure; indeed it cannot show whether the pressure-sensing element or other parts of the mechanism are broken or not. Proper calibration of barometers can only be accomplished by direct pneumatic connection to an appropriate pressure standard and adherence to well defined procedures that include pressure-cycling.

If your barometer's pressure sensing mechanism is one that is not significantly influenced by changes in the value of gravitational acceleration (ie most non-mercury barometers), it is functioning properly and you are interested in the pressure at the location of the barometer (rather that at sea level) there should be no need to make any adjustments when you move it to a different height.

Temperature effects - non-mercury barometers

Temperature affects all instruments to some degree but most non-mercury barometers are designed to be compensated, to a reasonable degree, for variations in temperature - provided that ambient temperature does not go outside prescribed limits or change too quickly.

Temperature and gravity corrections - Fortin and Kew Pattern mercury barometers

Each reading of a Fortin and Kew Pattern mercury barometer has to be corrected for the local value of gravitational acceleration, g, and the temperature of the instrument at the time of the measurement. Changes in the value of g cause a directly proportional change to the barometer's readings. Temperature variations affect mercury barometer readings because the materials from which they are made - generally steel, brass, mercury and glass - have very different coefficients of thermal expansion and there are no built-in compensating elements. (See How do I use a Fortin or Kew Pattern barometer?)

Methods for determining the local value of g (and also altitude) are shown on a separate page.

Standard conditions

Since 1955 mercury barometers made in the UK have been designed to measure pressures directly in their respective units when subjected to standard conditions, viz 0 °C and gravity 9.806 65 m s-2. The size of the corrections needed to take into account the actual temperature of the instrument and the acceleration due to gravity at the place of observation are not small and may amount to 0.5% of the reading (5 hPa) - see BS 2520 : 1983, Barometer conventions and tables, their application and use. This error is in addition to other effects caused by non-perfect operation of a barometer. Proper pressure readings depend on barometers being in good condition, properly mounted in a thermally stable environment and provided with adequate temperature measuring facilities. (See How do I use a Fortin or Kew Pattern barometer?)


The best and certainly the most accurate way to use a barometer is to set it up so that it measures its 'local' pressure - that is at its sensing level - properly; a pressure altitude correction should then be applied ONLY if a pressure value corresponding to one at mean sea level is required.

The full procedure would be to:

  • read the barometer (after setting its fiducial point in the case of a Fortin barometer)
  • add any correction that has been determined from a prior calibration of the barometer
  • apply temperature and gravity corrections (mercury barometers only)
  • IF REQUIRED, add an altitude correction to give a value of pressure equivalent to that at mean sea level

If you have a normal domestic aneroid barometer, its inherent potential accuracy will probably be quite low and the procedural detail for applying corrections might not make much difference in reality - it might be more convenient simply to off-set its readings by the magnitude of your local height correction - but don't say I said that.

Last Updated: 25 Mar 2010
Created: 9 Aug 2007