Watch our YouTube video How is temperature measured?
The most commonly used thermometer in the Bureau has been the liquid-in-glass thermometer, consisting of a reservoir of liquid (the bulb) and a slim capillary tube through which the liquid rises when warming and descends when cooling. A relative newcomer is the resistance temperature detector (RTD) or temperature probe, which has been used at some Bureau stations since the 1990s. This instrument relies on the fact that the electrical resistance of some materials changes with temperature in a predictable way, and so is a measure of the temperature of the device. The RTD is always used in an automatic weather station (AWS), and has been progressively replacing mercury-in-glass thermometers at other stations.
Construction of a maximum temperature thermometer
Maximum temperature has historically been measured with a mercury-in-glass thermometer which has a constriction in the neck of the thermometer tube. As the air temperature rises mercury is forced past the constriction. However, as the temperature falls the constriction prevents the mercury from returning to the bulb of the thermometer. The height of mercury in the tube remains at that reached at the hottest time of day. The thermometer is reset by gentle shaking.
At sites where there is a temperature probe the maximum temperature is derived from the RTD which is sampled by the AWS every second.
Alcohol-in-glass thermometers containing a moveable index are used to manually record minimum temperatures. When the temperature falls, the liquid and index move down the column, but when the temperature rises the index remains in the lowest position while the liquid expands up the tube. The position of the index indicates the lowest temperature reached since the last reset - which is achieved by tilting the thermometer, bulb end upwards. As for maximum temperature, at sites where there is a temperature probe the minimum temperature is derived from the 1 second observations of the AWS.
Construction of a minimum temperature thermometer
Air temperature (more specifically, dry bulb air temperature to differentiate from observations to measure humidity) is measured at various times of the day as part of the observing program to assist with weather forecasting. The instrument used is either a conventional mercury-in-glass thermometer, which does not have the constriction which is present in the maximum temperature thermometer, or a RTD.
The highest temperature over the 24 hours prior to observation at 9 am is recorded as the maximum temperature for the previous day. The lowest temperature for the 24 hours prior to 9 am is recorded as the minimum temperature for the day on which the observation was made. The thermometer is read to the nearest 0.1 degree Celsius. Air temperature is also measured at many stations at various times throughout the day; from one or two observations (at 9am and 3pm), to hourly and, for some locations, minute observations from an AWS.
Virtually all temperature data are sent electronically in ‘real-time’ to the Bureau for storing within the climate database. Most stations report all year round, but some stations only report for particular times of the year, such as during the fire weather season or the snow season. Daily temperatures may be missed due to observer illness, equipment malfunction or power failures (in the case of sites with a temperature probe), in which case the maximum or minimum temperature reported is the highest/lowest over the period since the last observation.
Where possible an observation station is established following common practice and guidelines, which enhances the ability to compare observations between stations and over time. Sites which measure temperature are on a schedule of annual and biannual visits. The purpose of these visits is to
Once the data arrive at the Bureau they proceed through a number of quality control processes to detect errors, which includes checking for:
The large volume of data associated with the more frequent observations (such as one minute data) limits the quality control of these observations to automatic processes, whereas some human interaction is involved in quality controlling maximum and minimum temperature data.
A simple way to obtain reasonably accurate temperature measurements is to purchase an electronic ‘weather station’ available from a variety of sources. These weather stations can provide electronic read-outs of many elements including temperature, humidity and atmospheric pressure. Another less expensive option is to purchase a maximum and minimum temperature thermometer and set it up as described next.
A thermometer should be mounted so that air can pass freely over the bulb or sensor while being protected from direct sunlight and rain. The Stevenson screen (pictured right) is an instrument enclosure that meets these requirements and is the standard at official Bureau sites.
The instrument enclosure should be located in the open, over a locally natural surface (for example grass, but not concrete), unaffected by nearby obstructions such as trees and buildings (at a distance of at least twice the height of obstructions, and preferably four times), and not placed in a hollow (which would render conditions unrepresentative). The screen door should face south in the southern hemisphere to avoid direct sunlight on the thermometers when making an observation. The thermometers should be placed in the shelter at about 1.2 m above ground level.
A compromise for those unable to construct or purchase a shelter would be to hang a thermometer from the branch of a tree with thick foliage, not too close to the trunk. It is important that the bulb is never in direct sunlight or close to buildings (which can radiate heat, particularly overnight).
Place a mixture of distilled water and ice (made from distilled water) into an insulated container along with the thermometer to test. Agitate the mixture until the temperature reading stabilizes, making sure that there is still a mix of ice and water. The mixture will maintain a temperature of 0°C until all the ice has melted, or until all the water has frozen. The difference between the temperature reading and zero degrees Celsius is the error at freezing point.
Where ever possible temperature measurements should be made at approximately the same time each day. Measurements made at 9am local time will be comparable with observations provided by the Bureau. When reading the temperature from a thermometer, always have your eye level with the surface of the liquid in the thermometer stem to avoid parallax error. In large thermometers it should be possible to read to the nearest tenth of a degree by estimating the position of the mercury surface between the engravings. However, resolving an observation to the nearest tenth of a degree does not imply that the temperature is necessarily accurate to within a tenth of a degree.
Stevenson Screen
Maximum and minimum temperature thermometers are placed inside a Stevenson Screen. The screen allows good air flow across the thermometers but prevents heating from direct sunlight.
Resistance temperature detector
The resistance temperature detector (RTD) is always used in an automatic weather station (AWS), and has been progressively replacing the liquid-in-glass thermometers at other stations.
Page updated: June 5, 2018