Air temperature seems like the most basic measurement imaginable, yet it’s one of the easiest to get badly wrong, and the majority of home weather stations report temperatures that are off — sometimes by several degrees — for reasons that have nothing to do with the quality of the sensor. Understanding how temperature is actually measured, and what corrupts it, explains why your station might disagree with the official figure down the road, and how to fix it.
The sensor itself in most modern stations is a thermistor — a small electronic component whose electrical resistance changes predictably with temperature. The station reads that resistance and converts it to a temperature value. Thermistors are accurate, responsive and cheap, and the technology is rarely the problem. Older instruments used liquid-in-glass thermometers, where a fluid expands up a calibrated tube, and these are still prized as reference instruments for their reliability, but the electronic sensor is what you’ll find in any consumer station today. The trouble lies not in measuring the sensor’s temperature, but in ensuring the sensor’s temperature actually matches the air’s.
That distinction is the whole ballgame. What you want is the temperature of the free air, in the shade, moving freely past the sensor. What you often get instead is the temperature of a sensor that’s been warmed by sunlight, by heat radiating off a nearby wall or patio, or by stagnant air trapped against the house. Direct sun is the worst offender: a sensor sitting in sunshine absorbs that radiation and heats well above the true air temperature, easily reading three, five, or more degrees too high. It’s no longer measuring the air at all — it’s measuring itself getting hot in the sun.
This is why proper temperature measurement requires a radiation shield, sometimes called a Stevenson screen in its classic louvred wooden form, or a stack of white plastic plates in modern compact versions. The shield does two jobs: it blocks direct and reflected sunlight from reaching the sensor, while its open, louvred design lets air flow freely through so the sensor stays in touch with the real air temperature. White surfaces reflect sunlight rather than absorbing it. The best shields are aspirated, meaning a small fan continuously draws air across the sensor, which eliminates the errors that creep in on calm, sunny days when air isn’t moving naturally. Without a shield, even a perfect sensor produces unreliable numbers.
Siting compounds or relieves the problem. The international standard places the temperature sensor about 1.25 to 2 metres above ground over a natural surface like grass, well away from buildings, paving and heat sources. Concrete and asphalt soak up heat during the day and radiate it back, so a sensor above a driveway reads warmer than one above a lawn. Walls do the same and also block air movement. A sensor mounted on the sunny side of a house, just above a patio, is in about the worst possible spot — and that, unfortunately, is exactly where convenience leads many people to put it.
So if your station consistently reads warmer than the official station nearby, especially on sunny afternoons, the culprit is almost certainly shielding and siting rather than a faulty sensor. The remedy is to house the sensor in a proper radiation shield, mount it at the right height over grass if you can, and keep it clear of walls, paving and the heat of the building. Do that, and a modest consumer thermometer will track the real air temperature closely. Skip it, and even an expensive sensor will mostly tell you how hot your wall got in the afternoon sun — which is a far less interesting thing to know.