How to Calibrate a Weather Station: Rain, Temp, Humidity

Before you calibrate anything, hear the caveat: Davis sensors leave the factory calibrated, and the most common "calibration" mistake is adjusting a healthy station to match an airport 15 km away. That usually means programming someone else's microclimate into your own instrument. Calibrate when you have evidence of drift — and here is how to gather that evidence properly for rain, humidity, and temperature.

Does your weather station actually need calibrating?

Usually not. Davis temperature, humidity, and rain sensors are calibrated at the factory, and a station that disagrees with the airport is far more often a siting or microclimate difference than a drifting sensor. Calibrate only after a controlled reference check confirms an error — or after a repair or part replacement that could have disturbed the sensor.

The trap is intuitive: your station reads 3 °C above the nearest official station on a sunny afternoon, so you dial in a −3 °C offset. But that gap was probably radiation error or a genuinely warmer site — we cover the diagnosis in why your station reads hotter than the airport — and now your station reads 3 °C low on every overcast day. Legitimate reasons to calibrate: a drip test shows your rain gauge consistently over- or under-counting, a salt test shows humidity off by more than a few percent, or a side-by-side check against a trusted thermometer shows a constant offset in all conditions.

How do you test a rain gauge with a drip test?

Pour a measured volume of water slowly through the funnel and compare against the math: expected tips = volume ÷ (collector area × depth per tip). The Vantage Pro2 and AeroCone collectors have a 214 cm² opening, and the metric spoon tips every 0.2 mm — so each tip is about 4.3 ml of water.

That gives clean target numbers for a 500 ml field check:

The word slowly is load-bearing. Dump 500 ml in at once and water overshoots the spoons mid-tip, so the gauge reads low even when healthy. Let it drip over half an hour or more — a bottle with a pinhole in the cap works well. Landing within a few percent of the target means the gauge is fine. Two hardware checks come first, though: put a spirit level across the funnel rim, because a tilted gauge fills its buckets unevenly, and clear any debris. The tipping bucket mechanism is simple enough that level, clean, and free-swinging covers most problems. Note the Vantage Vue's collector is smaller, so its per-tip volume differs — the same formula applies with its own collector area.

How does the salt test for humidity work?

A saturated solution of ordinary table salt holds the air in a sealed container at very close to 75% relative humidity, almost regardless of room temperature. Seal your humidity sensor (or a reference hygrometer) in an airtight box with a small open dish of salt-water slurry, wait several hours, and read. Within ±3–5% of 75% is acceptable.

Make the slurry wet sand, not brine: salt with just enough water that liquid pools slightly. Give it at least 6–8 hours to equilibrate, and keep the container out of sunlight and temperature swings. On a Davis station the outside temperature/humidity sensor is mounted inside the radiation shield and is not convenient to seal in a box, so the practical route is to salt-test a small reference hygrometer first, then sit it beside the station sensor in the shade of the shield and compare. If the station reads consistently high or low by more than a few percent against your verified reference, enter a humidity offset; if it is within tolerance, leave it alone.

How do you check temperature calibration?

Two trustworthy references exist: an ice bath and a known-good thermometer. An ice bath — a slushy, well-stirred mix of crushed ice and a little water — sits at 0.0 °C and is the classic check for any probe that can be immersed. The Davis outside sensor cannot be dunked, so use the side-by-side method instead.

Place a verified reference thermometer in the same radiation shield as the station's sensor, or right beside it in full shade, and compare over several hours — never in direct sun, where radiation error will swamp the comparison. Overcast, breezy conditions are ideal because both instruments then read true air temperature. If readings diverge only when the sun is out, that is a siting problem, not calibration. A genuine offset shows up as a constant gap in all conditions, day and night — that constant number is what you enter into the console, with its sign flipped.

Where do you enter calibration offsets?

Temperature and humidity offsets live in software; rain does not. The Vantage Pro2 console accepts temperature offsets in 0.1° steps over roughly a ±12.7° range, plus humidity offsets and wind-direction alignment — Davis documents the procedure in its calibration support article. WeatherLink offers equivalent settings for newer consoles.

The rain gauge is the odd one out: there is no software multiplier to correct a miscounting bucket, and that is by design — a gauge that fails the drip test has a physical problem. Re-level it, clean the funnel and spoons, check the mechanism swings freely, and re-test. Enter offsets sparingly and write down what you set and why; an undocumented offset is indistinguishable from drift a year later.

When should you replace a sensor instead?

Replace when drift returns after calibration or when the hardware is visibly worn. Humidity sensors age faster than anything else on a station — years of pollution and salt exposure degrade the sensing element — and on the Vantage Pro2 the combined temperature/humidity sensor is a replaceable part, so a failing one is a swap rather than a repair.

The same logic applies to a tipping bucket whose bearings have worn sticky: no amount of leveling fixes friction in the pivot. A humidity sensor that salt-tests 10% off, gets an offset, and is 10% off again next season is telling you its element is dying — chasing it with ever-larger offsets just delays the inevitable while logging bad data. A seasonal maintenance routine catches most of this early, when it is still a cleaning job rather than a parts order.

How does calibration show up in your data history?

Long-term charts are where drift becomes visible — and where your fixes are documented forever. On a Pro Weather site, a slow humidity decline over two years, or the step-change on the day you entered an offset, is plainly visible in the multi-year history. Since Pro Weather keeps every reading from the day you connect, you get a permanent before-and-after record of every calibration you make.

Common questions

How do I test my rain gauge accuracy?

Check it is level with a spirit level, then drip a measured volume of water slowly through the funnel — 500 ml over at least half an hour. On a Vantage Pro2 metric collector that should register about 23.4 mm. Within a few percent is a pass; pouring fast will read low even on a healthy gauge.

How much rain is one bucket tip?

Davis sells two versions: the metric collector tips at 0.2 mm of rain, the US collector at 0.01 inch (0.254 mm). In water terms, one tip on a Vantage Pro2's 214 cm² collector is roughly 4.3 ml (metric) or 5.4 ml (US). The console multiplies tips by that calibrated amount to compute totals.

Can you calibrate a Davis console?

Yes, for most readings. The console accepts offsets for inside and outside temperature (0.1° steps, up to about ±12.7°) and humidity, and a wind-direction alignment correction. Rainfall has no console calibration — the tipping bucket is factory-set, and errors there are fixed at the hardware: leveling, cleaning, or worn-part replacement.

How often should sensors be recalibrated?

Check rather than recalibrate: a yearly drip test for the rain gauge and an occasional salt or side-by-side test for humidity and temperature is plenty. Enter an offset only when a check fails consistently. Healthy Davis sensors hold factory calibration for years — frequent adjustment usually means a dying sensor or a siting problem.