A hygrometer is an instrument that measures humidity, the water vapour present in a gas. Most household hygrometers display relative humidity as a percentage, often beside temperature. Other instruments measure dew point or another moisture quantity. Understanding how hygrometers work explains why two rooms can contain similar amounts of water vapour but show different percentages, and why a sensor beside a cold window can disagree with one across the room.

Reading a hygrometer correctly means identifying the quantity, location, temperature and time represented by the number. A reading of 70% relative humidity does not mean that 70% of the air is water. Nor does a lower percentage necessarily prove that water has been removed. Temperature changes, sensor placement, calibration and response time can all influence the interpretation.

This guide explains humidity measurement from first principles, compares electronic sensors and wet-and-dry-bulb instruments, and distinguishes relative humidity from dew point and moisture in solid materials. Worked examples show how to interpret changes, compare sensors and plan a useful observation record. The purpose is to make the measurement understandable, not to diagnose a building, prescribe a medical humidity target or substitute a single display for professional investigation.

The 50-second explanation

Water vapour is the gaseous form of water. A hygrometer detects a physical effect associated with it: a sensing film changes electrical properties, evaporation cools a wet thermometer, or condensation appears on a deliberately cooled surface. The instrument converts that response into a stated humidity quantity.

Relative humidity compares the actual water-vapour pressure with the saturation vapour pressure at the same temperature. Because the saturation value changes strongly with temperature, the percentage can rise when air cools even without additional water vapour entering it.

The practical rule is simple: record temperature with relative humidity, and keep the sensor location and conditions in view. A humidity number becomes useful when it answers a defined question. Without that context, an accurate instrument can support an inaccurate story.

Water vapour is not the mist you can see

Water vapour itself is invisible. The visible cloud above warm water or near a cooling surface consists of tiny condensed droplets. Calling both things steam in everyday conversation is understandable, but an instrument explanation needs to distinguish gas from suspended liquid.

A room can therefore contain substantial water vapour while looking completely clear. Conversely, a brief visible mist near a source does not tell you the humidity throughout the room. The measurement location determines which part of the environment the instrument samples.

Evaporation transfers molecules from liquid water into the gas phase. Condensation transfers them back into liquid when conditions favour that change. In a closed system, these exchanges can approach equilibrium; in an ordinary room, ventilation, people, materials and temperature keep changing the situation.

Think about a wet towel. It loses water as evaporation proceeds, but the surrounding relative humidity depends on more than the towel. Room temperature, airflow and exchange with outside air all matter. The towel is a moisture source, not a complete description of the room’s moisture state.

Relative humidity: a ratio, not a percentage of the air

The numerator and denominator

Relative humidity can be written, in a simplified ordinary-air description, as RH = 100 multiplied by e divided by es(T). Here e is the actual water-vapour partial pressure and es(T) is the saturation vapour pressure at temperature T. Both pressures must use the same units.

Partial pressure describes the contribution associated with water vapour within the gas mixture. The saturation value refers to equilibrium with the appropriate condensed water phase at that temperature. This formulation is more precise than imagining air as a sponge with holes that fill with water.

At a fixed temperature, increasing water-vapour pressure increases RH. At a fixed water-vapour pressure, warming increases the saturation value and reduces RH. Cooling has the opposite effect until condensation or another exchange changes the water-vapour amount.

Singapore’s Meteorological Service explanation of observations distinguishes relative humidity from temperature and explains wet-and-dry-bulb measurement. Its definitions help separate an actual moisture change from a temperature-driven change in the percentage.

Why the sponge analogy eventually fails

The phrase warm air holds more water can be useful as a first memory aid, but it can suggest the wrong mechanism. Water molecules are not stored in empty compartments between air molecules. The relevant equilibrium vapour pressure changes with temperature.

This matters when reasoning about heating. Warming a room may lower its RH without removing the water vapour already present. Saying that the room has been dehumidified solely because the percentage fell confuses a changed denominator with a reduced moisture quantity.

A better sentence is: the air is farther from saturation at its new temperature. That statement explains the display without claiming an unmeasured removal of water. It also prepares the reader to understand why the percentage can rise again when the room cools.

Worked example: the same vapour pressure at two temperatures

Assume a simplified air sample has a water-vapour pressure of 1.17 kilopascals. Use approximate saturation pressures of 2.34 kilopascals at 20 degrees Celsius and 4.24 kilopascals at 30 degrees Celsius. These values are rounded for the calculation, not measurements of a particular room.

At 20 degrees, RH is 100 multiplied by 1.17/2.34, giving 50%. At 30 degrees, with the actual vapour pressure assumed unchanged, RH is approximately 100 multiplied by 1.17/4.24, giving 27.6%. The displayed percentage has fallen substantially.

No water removal was included in the assumptions. The example holds actual vapour pressure constant to isolate the temperature effect. A real sealed rigid container or changing room can involve additional pressure and exchange effects, so the assumptions should not be silently transferred to every situation.

The useful conclusion is not that heating always produces exactly this change. It is that comparing two RH percentages without their temperatures is incomplete. The calculation reveals what information is missing before an explanation about drying can be trusted.

For conversions involving several humidity quantities, Vaisala’s humidity calculator provides a technical reference. Supply the required temperature and pressure conditions rather than expecting one percentage to determine every other moisture quantity by itself.

Dew point answers a different question

Dew point is the temperature at which a gas sample would reach saturation during a specified cooling process, ordinarily considered at approximately constant pressure and before moisture is removed by condensation. It provides another way to describe the water-vapour condition.

A room’s dry-bulb temperature and dew point are not interchangeable. The dry-bulb value describes the current thermal condition. The dew point describes the temperature boundary associated with saturation for the sample under the stated conditions.

Consider a cold surface within a room. If the surface is below the local dew point, condensation may form there even when the room’s central RH display is well below 100%. The surface environment is colder than the sensor location, so the local saturation relationship differs.

That observation explains why a centre-of-room hygrometer cannot certify that every window recess, pipe or concealed surface is dry. It measures its own surroundings. A building investigation needs relevant surface temperatures and locations, not just a reassuring percentage displayed elsewhere.

A chilled-mirror instrument measures a condensation-related temperature directly using controlled cooling and optical detection. For that particular mechanism, continue to the chilled-mirror dew-point reading. Here it serves as a comparison with the broader family of hygrometers.

Humidity quantities that should not be mixed

Absolute humidity and mixing ratio

Absolute humidity commonly means water-vapour mass per volume of gas, often expressed in grams per cubic metre. Mixing ratio describes water mass relative to dry-air mass. These quantities have different denominators, so they respond differently when the sample expands, contracts or changes pressure.

A label saying moisture content is not enough. Ask whether the value refers to water per volume, water per dry-air mass, a mole fraction or a relative-humidity percentage. Units help identify the quantity before any comparison is attempted.

Suppose a report lists 12 grams per cubic metre and another lists 12 grams per kilogram. The matching numeral does not establish equal humidity. One denominator is volume and the other is mass. Temperature, pressure and the precise definition are needed for a meaningful conversion.

Moisture in a wall is another measurement

A wall moisture meter does not generally measure the same quantity as an air hygrometer. It may respond to electrical properties associated with the material, and its reading can depend on construction, salts and the instrument’s calibration model.

Do not treat a pin-type material meter and a room RH display as competing measurements of one variable. They can provide complementary evidence, but their numbers should not be compared directly as percentages without understanding what each percentage means.

The National Physical Laboratory’s humidity and moisture research treats measurement in gases and moisture in materials as related but distinct problems. The common word water does not make the measurement methods interchangeable.

How an electronic hygrometer detects water vapour

Many electronic humidity sensors contain a thin material layer whose electrical properties change as it exchanges moisture with surrounding air. A capacitive design detects a change in capacitance; other designs use different electrical responses. Temperature is measured or accounted for as part of the conversion.

The sensing layer must interact with the environment. This distinguishes a humidity sensor from a completely isolated electronic component. A protective filter can reduce contamination while allowing gas exchange, but a blocked opening or inappropriate enclosure can prevent representative measurement.

Vaisala describes a capacitive thin-film humidity probe with temperature measurement and environmental protection. The example illustrates the sensor, filter and shielding roles; it does not mean every inexpensive household unit has the same construction or performance.

The instrument converts a small electrical signal into a displayed value using calibration information. That information accounts for predictable relationships among signal, humidity and temperature. The resulting percentage is therefore not simply the raw capacitance written on a screen.

Electronic sensing makes compact displays and data logging possible. It also introduces questions about power, software, averaging and compensation. A clear digital readout is convenient, but clarity of typography is not evidence that the sensor has been placed correctly or remains calibrated.

Why temperature measurement belongs beside humidity measurement

A relative-humidity sensor must represent the thermal condition of the gas being assessed. If the sensing element is warmed by nearby electronics or direct sunlight, its local RH can differ from that of the undisturbed air even when the surrounding moisture source is unchanged.

A wall display near a lamp may therefore read lower than another device in a shaded position. The first question is whether the two sensors share comparable temperatures and airflow. Moving them together for a controlled comparison is more informative than immediately declaring the higher reading correct.

Protective radiation shields address this problem outdoors by reducing direct heating while allowing airflow. A sealed box protects against rain differently: it may create its own internal environment and delay gas exchange. The best protection must preserve the quantity the instrument is intended to sense.

Temperature compensation corrects known sensor behaviour; it cannot necessarily reconstruct the intended room condition when the probe has been installed in a different microenvironment. Software is not a substitute for representative placement.

How wet-and-dry-bulb measurement works

A psychrometer uses two temperature measurements. One thermometer measures the surrounding air temperature. The other has a wetted surface around its sensing region. Evaporation from that wet surface removes energy and can make it cooler than the dry thermometer.

The temperature difference depends on humidity, but also on airflow and the measurement conditions. In drier air, evaporation can produce a larger cooling effect. Near saturation, evaporative cooling becomes smaller. Suitable psychrometric relationships convert the readings into a humidity estimate.

The wet covering must be properly wetted and exposed to the intended airflow. A dry wick, contaminated material or inadequate ventilation changes the physical experiment. It is not enough to put two thermometers beside each other and call the lower reading a humidity value.

Pressure can also enter the calculation. The US National Weather Service humidity calculation resource distinguishes the needed inputs and uses station pressure where relevant. A remembered sea-level weather-map value should not automatically replace pressure at the measuring location.

For learners, the method is valuable because the mechanism is visible: liquid evaporates, the sensor cools, and two temperatures are compared. Its apparent simplicity should not hide the need for a defined procedure. Different airflow or wick conditions can produce different results without any change in the room’s actual humidity.

Chilled mirrors and reference measurements

A chilled-mirror hygrometer cools a surface while observing the onset or equilibrium of condensation. The measured surface temperature provides a dew-point or frost-point result, depending on the condensate and conditions. RH can then be calculated using additional temperature information.

This approach is conceptually different from a polymer film changing electrical properties. Comparing distinct measurement principles can help identify errors that would remain hidden if every instrument shared the same response mechanism.

NIST’s hygrometer calibration information describes exposing instruments to gas with known moisture conditions and controlling temperature when calibrating RH. The reference conditions matter as much as the display under test.

A reference instrument is not magic. Surface contamination, temperature measurement and the identity of condensed water or ice must still be controlled. The value of the method comes from a well-understood physical relationship supported by a careful measurement procedure.

At another level, NIST’s gravimetric hygrometer determines humidity through separated water and gas mass measurements. This illustrates how a familiar household percentage ultimately connects to more fundamental quantities and reference systems.

Calibration, checking and adjustment are different operations

Calibration establishes the relationship between an instrument’s indications and suitable reference values under stated conditions. A check asks whether performance remains acceptable for a purpose. Adjustment changes the instrument to reduce an observed error. These activities may occur together, but they are not synonyms.

Pressing a button that makes a sensor match another display is an adjustment. It becomes defensible only when the comparison is appropriate. Two unverified devices disagreeing does not identify which one should be changed, and averaging their displays does not create a reference standard.

A one-point check can identify an offset near that condition. It does not necessarily reveal nonlinearity across a broad humidity range or a temperature-dependent error. NPL’s relative-humidity calibration service describes checking multiple humidity and temperature points for this reason.

For ordinary learning, a careful comparison under shared conditions can be useful without pretending to be a formal calibration. State its limitation: these devices agreed within a certain amount in this room after settling. Do not transform that statement into a certificate of accuracy in all climates.

Breathing on a sensor may demonstrate that it responds, but it is not a calibrated humidity test. Breath changes temperature and moisture together and can expose the sensor to condensation. A moving display proves responsiveness, not correctness.

Worked example: percentage points versus percentage error

Suppose an instrument indicates 60% RH and a specification is expressed as plus or minus 3 percentage points of RH under the relevant conditions. The corresponding interval around the indication is 57% to 63% RH. It is not 58.2% to 61.8%.

The narrower interval would represent plus or minus 3% of the numerical reading, because 3% of 60 is 1.8. These are different statements. A specification written as percent RH often refers to absolute percentage points, but its precise wording and conditions should be checked.

Now imagine comparing two instruments whose uncertainty ranges overlap. A small displayed difference does not necessarily prove that either is faulty. Conversely, overlap alone does not establish that both are correct; they may share a bias or experience different local conditions.

Write units and uncertainty language explicitly. A phrase such as about three percentage points apart is less ambiguous than three percent wrong. The purpose of the calculation is to prevent the notation from creating certainty that the measurements do not support.

Response time is not the same as display refresh

An instrument may update its screen every second while the sensing layer takes much longer to approach a new humidity condition. The electronic display interval and the physical response time are different parts of the measurement chain.

Moving a sensor from a dry room to a humid one does not guarantee an immediate final value. Moisture must reach the sensing material, the sensor must respond, temperature may need to settle, and software may average the result. A protective filter can add another delay.

A simple hypothetical model illustrates the effect. Suppose a sensor’s response is described approximately by a 20-second time constant after a sudden environmental change. After one time constant, it has covered about 63% of the total change; after three, about 95%. Those figures describe the model, not a universal household hygrometer specification.

Waiting only five seconds would understate the final difference. Waiting longer while the environment itself continues changing creates a different problem: the sensor is following a moving target. Record both the observation timing and the conditions rather than assuming one universal settling period.

For a room survey, use a repeatable placement and settling procedure appropriate to the instrument. A rushed tour in which each room receives a different waiting time can rank rooms by sensor history instead of humidity.

Placement: the room is not one perfectly mixed box

Humidity and temperature can vary near windows, supply vents, wet objects and exterior walls. A sensor next to a source answers a local question. A sensor in a more representative occupied area answers a broader question. Neither placement is inherently wrong; confusion arises when the local reading is presented as the whole room.

Avoid unnecessary direct heat, splashes and contact with damp surfaces. Leave the sensing opening exposed as intended. Placing a hygrometer in a tightly closed drawer to protect it changes the environment being measured, even if the drawer is in the room of interest.

For comparisons, record height, distance from obvious sources and whether doors or ventilation conditions changed. Exact architectural surveying is not necessary for every diary, but enough description should remain for someone to repeat the observation.

Two sensors placed together can help test consistency. Keep them close enough to share conditions without blocking each other’s ventilation or warming each other. After suitable settling, exchange their positions. If the difference follows the location rather than the device, a spatial effect becomes more plausible.

That exchange is a small but powerful experiment. It separates two hypotheses that a single side-by-side photograph cannot resolve: different instruments and different surroundings.

Worked example: a cold corner and a warmer centre

Assume the water-vapour pressure is approximately the same across a small room, but one corner is cooler than the centre. The cooler location has a lower saturation vapour pressure, so its RH can be higher. A central reading does not capture that local thermal effect automatically.

For an illustrative calculation, let actual vapour pressure be 1.40 kilopascals. With an assumed saturation value of 2.80 kilopascals at one location, RH is 50%. At a cooler location where the assumed saturation value is 2.00 kilopascals, RH is 70%.

The example uses supplied rounded pressures to isolate the ratio. It does not assign those values to a particular building or claim that vapour pressure must be uniform in every room. Real sources, airflow and material exchange can also produce moisture differences.

A useful follow-up measures local temperature and observes the surface condition. A poor follow-up simply relocates the sensor until it displays the preferred number. Measurement should reveal the environment, not hide a condition by choosing a more flattering position.

Logging: a curve is a processed description

A spot reading answers what the instrument indicated at one time. A logger reveals duration, timing and variation. That distinction matters when a room becomes humid briefly after an activity and then returns toward its earlier condition.

Record the sampling interval and whether points are instantaneous values or averages. A short spike may appear clearly in a one-minute log and almost disappear in an hourly average. Neither display must be wrong; they describe different time scales.

Graphs also depend on their axes. A change from 58% to 61% can look dramatic on a narrow vertical scale and modest on a full 0-to-100 scale. Read the numerical difference before interpreting the visual steepness.

Keep temperature beside RH on the same time record, even if plotted separately. A rising RH trace accompanied by falling temperature invites a different explanation from rising RH at stable temperature. The paired record does not answer everything, but it prevents an avoidable missing-variable error.

Preserve gaps rather than inventing data. A line joining observations before and after a power outage is a drawing convention, not evidence that humidity changed linearly throughout the missing interval.

A practical observation plan for learners

Choose a safe indoor location and a functioning hygrometer whose instructions can be checked. Identify the temperature unit, RH display and any minimum, maximum or memory modes. A stored maximum is not the present reading, even when both appear on the same screen.

For several days, record observations at consistent times. Note ordinary activities that might affect the environment, such as opening windows or drying clothes, without deliberately creating damp conditions. Keep the device away from water and do not dismantle its sensing element.

Write observations before explanations. The display rose from 55% to 63% while temperature fell is a record. The room gained water is one possible interpretation, but not the only one. An explanation should account for both temperature and moisture pathways.

Compare a short period with a longer one. Did the reading recover after an activity? Did the same timing recur on another day? Was the sensor moved? A pattern that survives these checks deserves more confidence than a dramatic isolated number.

Finish by identifying a limitation. The sensor monitored one location, not every surface. The comparison lacked a calibrated reference. The sampling interval might have missed short changes. These statements improve the usefulness of the result because they show exactly what was and was not established.

Troubleshooting without guessing from one number

Two displays disagree

Check quantity, temperature, location, timing and memory mode first. Then allow comparable settling and repeat the comparison. A disagreement that persists with devices exchanged between positions deserves an instrument-focused check, but it still does not identify the accurate device without a suitable reference.

The reading changes after heating

Ask whether temperature changed faster than moisture could leave or enter. A lower RH after heating does not alone establish water removal. Review the simultaneous temperature trace and the sensor’s exposure to direct heat before attributing the entire change to drying.

The display stays near its upper limit

Consider actual high humidity, condensation on the sensor, a wet filter, contamination or a sensor fault. Do not wipe the sensing element or apply heat arbitrarily. Follow the manufacturer’s recovery and maintenance instructions, then verify operation before relying on important readings.

The room smells damp despite a moderate reading

A single central measurement cannot locate hidden moisture, explain an odour or assess all surfaces. Different locations and materials may be involved. Persistent building concerns warrant appropriate inspection rather than treating the display as a complete clearance certificate.

The last digit keeps moving

Small fluctuations may reflect actual variation, electronic noise or rounding. Identify the magnitude and time scale before interpreting every change. Filtering can improve readability, but it can also conceal short events; choose it according to the question rather than aesthetics.

Three comparisons that answer different questions

Has this particular room changed?

Keep the device in the same place and compare observations with temperature recorded. This design makes the sensor a consistent observer of one location. It does not remove calibration error, but a stable offset may matter less when the question concerns a substantial change over time rather than the absolute value.

Be careful when the instrument itself changes. Replacing a battery, moving the enclosure or switching from one device to another can create a discontinuity. Mark that event in the log. Otherwise a maintenance change may later be interpreted as a change in the building.

Are two locations different at the same time?

Two simultaneous sensors avoid the delay of carrying one instrument between rooms, but they introduce a possible difference between sensors. A useful comparison first places both together, then separates them, and finally exchanges their positions when practical. That sequence helps distinguish location effects from persistent instrument differences.

The experiment still needs a clear definition of location. A sensor at a cool window and one beside a warm appliance are not automatically representative samples of two whole rooms. Write the actual placement so the conclusion remains as specific as the observations.

Has an intervention changed moisture rather than just temperature?

Suppose a fan is switched on and the RH reading changes. The fan may improve mixing, alter the sensor’s thermal condition or change evaporation from nearby wet material. The observed percentage alone does not identify which pathway dominated. Record temperature and the relevant airflow arrangement before attributing the result to water removal.

Likewise, opening a window does not guarantee that the room becomes drier in every sense. Outdoor and indoor temperatures may differ, and their RH percentages cannot be compared as though they were water masses. Use comparable quantities and conditions, or describe the result narrowly as an observed change at the indoor sensor.

Turning sensor readings into useful summaries

A daily maximum can reveal that a threshold was crossed, but it does not show duration. A room briefly reaching a high value during an activity is different from one remaining near that value for many hours. Record time spent in a range when duration is relevant, while avoiding unsupported claims about a universal safe threshold.

A daily average hides a different kind of information. Imagine a hypothetical record alternating equally between 40% and 80% RH. Its average is 60%. Another room steady at 60% has the same average but a very different pattern. The arithmetic is correct in both cases; the average simply answers less than the full trace.

Do not average unlike quantities without a purpose. Combining a basement sensor’s RH with a warm upstairs sensor’s RH produces a numerical mean, but that mean is not necessarily a representative humidity condition anywhere. A building summary should preserve location and temperature differences rather than flattening them into one reassuring number.

A useful concise report states the observation period, sensor position, typical range and important departures. It can then separate possible causes from established ones. This makes the record easier to act on without pretending that a descriptive summary is already a diagnosis.

Reading a specification with the actual task in mind

Start with the stated measuring range and environmental limits. Then inspect uncertainty or accuracy language, response time, temperature dependence and whether condensation is permitted. A sensor intended for ordinary indoor observations may be inappropriate inside a wet process even though both environments can be described with humidity percentages.

Look for the conditions attached to the best advertised number. Performance near the centre of a range may differ from performance near its limits. A typical value is not always a guaranteed worst-case bound. A factory calibration statement does not mean the instrument cannot drift or become contaminated later.

Finally consider access for maintenance and comparison. A permanently hidden sensor may provide a convenient graph while becoming difficult to inspect after a leak or renovation. Good installation includes a way to verify the measurement later. The lifetime usefulness of a hygrometer depends on preserving confidence in the sensing path, not merely keeping the screen switched on.

Frequently asked questions

What is the difference between a hygrometer and a barometer?

A hygrometer measures humidity. A barometer measures atmospheric pressure. A weather station may contain both sensing functions, but one quantity does not directly substitute for the other. For pressure interpretation, use the barometer guide.

Does 100% RH mean the air is entirely water?

No. It means the water-vapour pressure is at the relevant saturation value for the temperature and stated conditions. The air still contains its other gases. Relative humidity is a comparison with saturation, not the fraction of the atmosphere made of water.

Can RH fall without a dehumidifier removing water?

Yes. Warming can lower RH by increasing the saturation value. This is why temperature must accompany comparisons. An actual dehumidification claim needs evidence about moisture removal or a suitable humidity quantity under comparable conditions.

Is a digital hygrometer always more accurate?

No. Accuracy depends on the sensing system, calibration, environment and maintenance. A digital display can be easy to read while measuring an unrepresentative location. Choose an instrument by the required performance and operating conditions, not merely the display format.

Why should I wait after moving the sensor?

Both humidity response and temperature equilibration take time. The sensor can temporarily reflect its previous environment. Use the manufacturer’s guidance and a repeatable procedure instead of assigning a universal waiting time to every instrument.

Does one correct comparison prove the whole range is accurate?

No. An instrument can agree near one humidity level yet differ elsewhere or at another temperature. Multiple reference conditions reveal errors that a single-point comparison cannot. A simple household check should be described as limited rather than presented as a full calibration.

Can a hygrometer detect a water leak?

It may reveal a humidity change consistent with several causes, including a moisture source, but it does not directly locate or identify a leak. Combine relevant observations and obtain appropriate investigation when the consequences matter.

What should a useful record include?

Include time, location, RH, temperature, instrument identity, reference or memory mode, and relevant changes in surroundings. Keep original readings separate from calculations and explanations. That record lets another person understand why a conclusion was drawn and how to check it.

The bigger picture: humidity is a relationship

A hygrometer makes an invisible condition readable, but its output remains tied to definitions and surroundings. Relative humidity is especially instructive because its denominator changes with temperature. A falling number can therefore describe a changing relationship without proving that material has left the room.

The reliable chain is physical response, calibrated indication, representative placement and careful interpretation. Each link can fail independently. Better digits do not solve poor placement, and careful placement does not repair a damaged sensor.

Learn to ask which water quantity was measured, at what temperature, where and over what interval. Those questions turn a household display into a useful scientific instrument and build habits that transfer to every measurement whose meaning depends on context.

Useful routes

Explore condensation on a cold can for a visible dew-point connection, the chilled-mirror manual for a specific reference technique, and barometers for pressure rather than humidity. The numerical cases in this article are illustrative models, not observations or building assessments.