Field NotesComparisonsNo. 18
How to choose a hygrometer for plants and check its accuracy
A humidity meter reports the air at its own sensor, at its own temperature. Here is what the specs mean, where to put it, and how a spoonful of salt tells you whether it reads true.

The short answer Choose a digital hygrometer that reads temperature and humidity at the same sensor and states its accuracy, with the conditions it applies to. Good sensor chips are rated about ±1 to ±3% RH at 25 °C and ±0.1 to ±0.5 °C. Mount it at leaf height, out of direct light, then check it in a sealed container with damp table salt, which holds about 75% RH (75.3% at 25 °C).
A hygrometer reports the humidity at its sensor, at the sensor’s own temperature. Whether that matches the air around your leaves depends on the sensor, where it sits, and whether it has been tested. You can test it yourself with salt and a sealed box.
What a hygrometer measures, and why temperature counts too
Relative humidity (RH) is the water vapor in the air as a share of the most the air could hold at that temperature. Warm the same air and RH drops, though no water has left. Plants respond to both, and vapor pressure deficit (VPD) combines them. The VPD note explains it, and the VPD calculator does the arithmetic.
So the thermometer matters as much as the humidity sensor. The World Meteorological Organization (WMO) instruments guide quantifies it: when RH and temperature are combined into a derived value such as dew point, a 0.1 °C error near 20 °C has the same effect as an RH error of 0.6% of the reading.
Example: for dew point near 20 °C, a ±0.5 °C thermometer error weighs about the same as ±3% of the RH reading (5 × 0.6%), or roughly ±1.8 points at 60% RH, on top of the humidity sensor’s own tolerance.
Hygrometer types and their typical accuracy
Digital (often capacitive). Most digital meters use a humidity chip, and many chips are capacitive. Sensirion’s humidity chips “are based on the capacitive measurement principle”: a dielectric polymer absorbs or releases water as humidity changes, which shifts its dielectric constant, and the chip measures that. Sensirion rates its SHT40 at ±1.8% RH typical and its SHT45 at ±1.0% typical, both at 25 °C. The SHT40’s listed maximum is ±3.5% RH (10 to 90% RH), the safer number to plan around.
Dial (mechanical). A dial moves its needle with a material that changes length as it takes up water; WMO names hair and synthetic fibers, with levers amplifying the motion. A good hair hygrograph “in perfect condition” reads to ±3% RH at moderate temperatures, worse in the cold. Between 0 and 30 °C and 20 to 80% RH, it should show 90% of a sudden change within about 3 minutes. Dust can throw it off by perhaps as much as 15% RH. Treat any dial as approximate until you’ve tested it.
Psychrometer. It pairs a dry thermometer with one in a wet wick and, in WMO’s words, “measures evaporative cooling of a wet surface.” It needs air moving past the bulbs at 2.2 m/s or more, distilled water and a clean wick, and it’s less accurate in dry air. WMO says sling psychrometers “tend to have significant uncertainty,” so use one as a cross-check.
| Type | How it senses | Accuracy (example or reference) | Watch for |
|---|---|---|---|
| Digital, capacitive | Polymer absorbs water, changing its dielectric constant | ±1.0 to ±1.8% RH typical at 25 °C (Sensirion SHT45, SHT40) | Self-heating, long humid spells |
| Dial (hair or fiber) | Element changes length; levers move a needle | ±3% RH, good hair hygrograph in perfect condition (WMO) | Slow response, dust, cold |
| Psychrometer | Evaporative cooling of a wet wick | 0.5 °C wet-bulb error ≈ 4% RH at 20 °C, 50% RH (WMO) | Airflow, wick care; worse in dry air |
Reading a humidity spec sheet
Where the accuracy applies. Bosch rates its BME280, a combined humidity, pressure and temperature chip, at ±3% RH from 20 to 80% RH at 25 °C. Datasheet figures describe the bare chip; Sensirion says its specs “hold for the stand-alone sensing element.” The sensors comparison covers what a finished device adds.
High humidity. This matters in domes, terrariums and tents. Sensirion recommends 20 to 80% RH (5 to 60 °C) for the SHT4x; long spells above can offset readings temporarily (+3% RH after 60 hours above 80%, in its example), and the sensor recovers by itself once back in that range. WMO links high or condensing humidity to long-term drift in some capacitive sensors; upward drift means over-reading, such as showing 100% at 95%. The BME280 is rated for non-condensing conditions only. The SHT4x is “fully functional in a condensing environment,” but it stops responding to RH while liquid water sits on it. Its heater removes condensed water; specs don’t hold while the heater runs.
Hysteresis. A sensor can read differently depending on whether it came from wetter or drier air. The SHT4x lists 0.8% RH; the BME280 lists ±1% RH, and its ±3% tolerance includes hysteresis.
Response time. The SHT4x takes 4 seconds and the BME280 1 second to reach 63% of a step change, both in air moving at about 1 m/s. Bosch says that airflow is needed to see its stated response, so a sensor in still air lags.
Long-term drift. Typical drift is under 0.2% RH a year for the SHT4x in its normal range; Bosch lists 0.5% RH a year for the BME280. Sensirion warns drift may be higher around solvent vapors, out-gassing tapes, adhesives and packaging.
Where to put it
- Measure where the leaves are: at canopy height, among the plants.
- Keep it out of direct light. Radiant heating can warm a sensor above air temperature, which WMO says “can potentially give a falsely low reading of relative humidity.”
- Watch for self-heating. Bosch says the BME280’s temperature reading “is typically above ambient temperature” because of board heat and self-heating. A warm sensor reads RH low; a probe on a cable keeps it away from the electronics.
- Let air move past it. Response specs assume about 1 m/s; a sensor in a dead corner lags.
- Shield it in a greenhouse or outdoors. WMO says to mount hygrometer probes inside a thermometer screen.
- Avoid heat and moisture sources. WMO warns against nearby heat and water vapor sources, the observer included: skip humidifiers, misters and heaters, and don’t breathe on it.
Logged readings vs. spot readings
RH swings with lights, heating and watering, so one reading can mislead. A logger records the swings, including humid overnight stretches.
Logs also play to what digital sensors do best. Sensirion lists SHT4x repeatability at 0.08 to 0.25% RH (three standard deviations, depending on mode), far tighter than the SHT40’s ±1.8% typical accuracy. A fixed sensor compared with itself shows real changes even if it reads a point or two off.
Handheld meters suit spot checks and cross-checks. Let each reading settle, especially on a dial, and hold the meter away from your body.
How to check accuracy with the salt test
A saturated salt solution, with undissolved salt left over, holds the air above it at a humidity that depends only on temperature. WMO lists sodium chloride at 75.3% RH at 25 °C. Lewis Greenspan’s 1977 compilation for the National Bureau of Standards (now NIST) gives:
- 68°F (20 °C): 75.47 ± 0.14% RH
- 77°F (25 °C): 75.29 ± 0.12% RH
- 86°F (30 °C): 75.09 ± 0.11% RH
Those values are for pure sodium chloride, so a kitchen-salt test is a check, not a lab calibration. You need table salt, distilled water, a small, wide cup and the smallest airtight container that fits your hygrometer.
- Make a damp slurry. Vaisala’s salt-calibrator manual uses 20 g of salt (about 15 ml, or 1 tablespoon) to 10 ml (2 teaspoons) of water, ending at 60 to 90% undissolved salt. Aim for wet sand, not brine.
- Seal the cup and hygrometer together. Keep salt off the sensor. WMO wants a salt surface large compared with the enclosed air, so equilibrium comes quickly.
- Hold the temperature steady. WMO says to keep the vessel at a constant, measured temperature. Pick a spot away from sun, heaters and drafts.
- Wait about a day. Vaisala allows about 24 hours for a fresh solution to reach equilibrium. Read the display through the container, and confirm the number has stopped changing.
- Note the temperature and find the matching reference above.
Even with table salt, follow WMO’s general advice for salt solutions: keep it off your skin and out of your eyes and mouth.
What to do with an offset
Subtract the reference from your reading. Example: at 68°F (20 °C) the reference is 75.47%, so a hygrometer showing 72% reads about 3.5 points low.
- Compare it with the spec. A BME280-based meter showing 73% at 25 °C is within Bosch’s ±3% (75.29 − 73 = 2.3).
- Apply the correction. Enter it in the meter’s offset setting if it has one; otherwise add it to readings in your log.
- Know its limits. The test checks one point, near 75%. WMO notes calibration error can appear as non-linearity, so the offset may not hold at 40% or 95%. Calibration kits add points with other salts, such as lithium chloride (11.1% RH at 25 °C, per WMO).
- Retest. WMO says field humidity instruments need regular calibration. After a long spell above 80% RH, give the sensor time in room air first. If the offset keeps growing, trust its trends, not its absolute numbers, or replace it.
A buying checklist
- Temperature and RH from the same sensor, both with stated accuracy.
- Tolerances with the RH range and temperature they apply to, ideally typical and maximum.
- Hysteresis, response time and drift listed, or a named sensor chip whose datasheet lists them.
- Rated for condensing conditions, or a heater, if it lives in a dome or tent.
- A probe on a cable, so the sensor sits at canopy height in shade.
- Logging with min/max and data export.
- An offset or calibration setting.
- A ventilated radiation shield for greenhouse or outdoor use.
Sources
- Humidity Fixed Points of Binary Saturated Aqueous SolutionsJournal of Research of the National Bureau of Standards, Section A, Vol. 81A, No. 1 (Greenspan, 1977), hosted by NIST
- Chapter 4. Measurement of humidityWorld Meteorological Organization, Guide to Instruments and Methods of Observation (WMO-No. 8), Volume I, preliminary 2018 edition
- Datasheet SHT4x: 4th Gen. Relative Humidity and Temperature SensorSensirion (datasheet version 6.5, April 2024)
- Humidity sensorsSensirion (product brochure)
- BME280 Data SheetBosch Sensortec (BST-BME280-DS001, revision 1.24, February 2024)
- Preparing solutions - HMK15Vaisala (HMK15 User Guide, M210185EN-E)
General information, not professional advice for a specific plant or crop. Spot something wrong? Email info@terrathink.io.


