How often should a weather station be calibrated?
There is no universal calibration interval. Set it from the measurement requirement, consequence of error, operating environment and evidence from previous results, then treat each parameter according to how its sensing technology behaves.
There is no universal calibration interval for every weather station. The appropriate interval depends on the measurement requirement, consequence of error, operating environment, sensing technology, quality requirements and any regulatory, contractual or customer requirement.
Ask how often a weather station should be calibrated and it is tempting to answer with a number. There is not one.
A weather station is a collection of different measurement technologies sharing an installation. Wind, humidity, temperature, pressure and rainfall do not necessarily drift at the same rate or fail in the same way. The appropriate interval therefore starts with the measurement requirement, not the calendar.
Requirement → consequence → environment → sensor behaviour → evidence
If a standard, contract or regulator specifies an interval, that requirement takes precedence. If it does not, the organisation needs a defensible basis for choosing one, and that basis should become stronger with every calibration result.
The interval belongs to the measurement, not the weather station
Calling something a weather station can hide the fact that it contains several different sensing technologies. A station may include wind speed and direction, temperature, relative humidity, pressure, rainfall, and solar radiation or other application-specific measurements.
Each parameter has its own sensing mechanism, environmental exposure and potential failure modes. A single calibration interval for the entire station may be convenient administratively, but it does not necessarily reflect how the individual measurements behave.
When a fixed interval already exists
If a regulator, standard, contract, customer specification or quality system defines the interval, follow that requirement.
When no fixed interval exists
Build the interval from the application and the evidence. The goal is not to find a universally correct number. The goal is to establish an interval that can be justified.
The consequence of being wrong matters more than the age of the sensor
A weather measurement used for general observation has a different consequence from one used to support an operational safety decision, trigger a crane stoppage, support aviation activity, feed a regulatory or contractual dataset, control an industrial process or provide long-term environmental evidence.
None of those applications automatically requires a particular interval. The principle is simpler than that: the more consequential the measurement, the stronger the evidence needed to justify the interval.
Build the interval from evidence, not habit
- Check whether a regulator, standard, contract, customer or internal quality system specifies the interval.
- Ask what happens if the measurement is outside tolerance.
- Salt, dust, icing, fouling, contamination, temperature exposure, physical access and installation conditions.
- Review successive calibration results. Where an instrument repeatedly remains comfortably within tolerance, there may be evidence to support extending an internally controlled interval, where the governing requirement permits it. A marginal result is evidence for investigation and potentially a shorter interval.
Your certificate will not set the interval for you
A certificate from an accredited calibration laboratory will not tell you when to recalibrate. ISO/IEC 17025:2017, clause 7.8.4.3, does not allow a calibration certificate or label to recommend a calibration interval unless that has been agreed with the customer. The interval belongs to the instrument owner, which is why it has to be built from your own requirement, consequence and evidence.
“We always calibrate annually” is not much of a justification
An annual interval may be appropriate for a particular application. But the fact that an organisation has always used twelve months does not, by itself, establish that twelve months is technically necessary. Equally, a long history of stable results does not permit an interval change if an external requirement specifies the frequency.
The defensible position is to know why the interval exists, then use evidence to maintain or challenge it.
A weather station is several measurement problems sharing one mast
| Parameter | Typical behaviour to consider | Practical implication |
|---|---|---|
| Ultrasonic wind | Fixed sensing geometry and timing-based measurement, with no moving mechanical element | Verification and production method matter differently from mechanical sensors |
| Relative humidity | Sensing element can be more susceptible to drift and contamination | Often deserves particular attention when setting the interval |
| Temperature | Generally stable, but installation and radiation shielding affect the measurement | Check the probe and the physical installation |
| Barometric pressure | Generally stable and comparatively easy to compare against another source | Useful for ongoing plausibility checks |
| Tipping-bucket rainfall | Mechanical movement and potential blockage | Inspection and condition checks are important |
| Optical rainfall | No moving mechanical measurement element | Verify performance and keep the optical path clean |
Humidity can become the parameter that drives the interval where drift and contamination are significant. That is an observation about a common case, not a rule for every station. Equally, an ultrasonic wind sensor is not exempt from calibration. Its verification and production-control approach differs from a sensor whose output depends on mechanical movement or on an element that drifts.
How accuracy is established at Gill
The answer to how often to calibrate depends partly on how the instrument’s accuracy was established in the first place, and Gill does not use one process for every product.
Some Gill products are not individually calibrated or adjusted, because their accuracy is established through controlled manufacturing processes and the matching and accuracy of the sensing transducers.
GenCal: production verification
Gill’s GenCal process takes every fifth unit to a highly accurate wind tunnel to check that production remains within specification. This is production verification. It is not individual calibration of every fifth unit, and it is not a statistical proof of the units in between. It establishes that the manufacturing process is holding its specification.
When every unit is tested
For higher-value products such as WindMaster, each unit is individually adjusted during manufacture and every unit is tested in the wind tunnel.
Calibrated is not the only way an instrument can be made accurate.
Accuracy can be established through a combination of sensor design, transducer matching, controlled manufacture, adjustment where appropriate and production verification. One method is not universally superior to another. They answer different questions.
Calibration, adjustment and verification are not the same thing
Calibration
Establishing, under specified conditions, the relationship between the values given by a measurement standard and the instrument’s indications, each with its measurement uncertainty, and documenting the result. Calibration on its own does not change the instrument.
Adjustment
Operations carried out on the instrument so that it gives prescribed indications for given values of the quantity being measured. Adjustment is not calibration, and an adjusted instrument normally needs calibrating again. An instrument can be calibrated without being adjusted.
Verification
Providing objective evidence that the instrument meets specified requirements, such as a stated tolerance. A calibration result is often that evidence, but calibration and verification are not the same thing.
Production verification
Checking representative production output to establish that the manufacturing process remains within specification.
A production test can verify that a manufacturing process is holding its specification. A field check can verify that an installed instrument remains plausible. Neither is a laboratory calibration.
A calibration certificate is evidence of a comparison, not a clean bill of health
A certificate provides evidence of how an instrument compared with a reference under the conditions covered by the calibration. It does not automatically prove correct installation, correct siting, correct orientation, correct configuration, correct data transmission, suitability for the application or future performance. Calibration answers one question. It does not answer every question about the measurement system.
Traceable, factory tested and UKAS-accredited are different statements
Factory testing is evidence generated during manufacture.
Traceable calibration is a calibration whose result can be related to a stated reference through a documented, unbroken chain of calibrations, each contributing to the measurement uncertainty.
UKAS-accredited calibration is calibration performed by a laboratory operating under an accredited scope. The accreditation relates to the laboratory’s competence and the activities within that scope. It does not mean every instrument calibrated there is accurate in every application.
Labcal is a UKAS-accredited laboratory within Gill Group, UKAS Lab 0625, operating to ISO/IEC 17025:2017. Its scope includes air velocity, temperature and humidity, pressure, gas flow and liquid flow, and it supports equipment from manufacturers other than Gill as well as Gill equipment. Labcal does not calibrate every Gill instrument, and not every Gill weather sensor falls within its scope.
The most useful check may happen between calibrations
Calibration gives periodic evidence. Operational monitoring can provide earlier warning.
Status information
Where the instrument provides status information, log it alongside the measurement data. WindSonic status codes are available on every output. A change in the frequency of flagged or abnormal outputs can be an early indication that deserves investigation. Status codes do not diagnose every failure.
Compare with a nearby reference
Pressure and temperature can sometimes be compared with another trusted station or reference. This is a plausibility check, not a substitute for calibration.
Inspect the physical installation
Keep radiation shields and optical paths clean where applicable, and photograph the installation during site visits. That provides evidence of physical condition over time.
Preserve configuration
Keep the configuration file so a replacement can be configured consistently. This supports continuity of the measurement record.
Between calibrations, look for change rather than waiting for a certificate
A comparison with another station does not establish calibration. A clean optical path does not establish calibration. A status code does not establish calibration. All three can provide useful evidence that tells the operator when further investigation is warranted.
Calibration has an operational cost. Plan for the gap.
The instrument has to be removed from service for laboratory calibration. That creates a measurement gap unless the site has a swap unit, a planned outage, another valid source or an agreed data-quality procedure. Not every site needs a spare, but every site needs a plan.
Where access is expensive, align calibration with other planned maintenance or site work. The aim is not simply to minimise visits. It is to minimise unnecessary operational disruption while preserving the required measurement assurance.
Decide what happens to the measurement record before you remove the sensor
If a weather station supports a long-running dataset, the calibration event itself becomes part of the data-management plan. Record the timestamp of removal, the replacement instrument, configuration continuity, the data gap, instrument identity, the calibration certificate and the return-to-service date.
So how often should a weather station be calibrated?
There is no single interval that applies to every weather station or every parameter.
If a regulator, standard, contract or quality system specifies the interval, follow it. If it does not, set the interval from the consequence of measurement error, the operating environment, the sensor technology, historical calibration results and the organisation’s quality requirements. Then review the interval when the evidence changes.
What should make you shorten the interval?
- A calibration result approaches the tolerance boundary
- A calibration result falls outside tolerance
- Environmental exposure becomes harsher
- Contamination increases
- Sensor behaviour changes
- Status information indicates an emerging issue
- Application consequence increases
- Quality or regulatory requirements change
A marginal result does not automatically mean the interval must be shortened. It may justify investigation and potentially a shorter interval.
When can the evidence support a longer interval?
Where no external requirement fixes the interval, repeated results that remain comfortably within tolerance can provide evidence for reviewing whether the existing interval is unnecessarily conservative. The decision still belongs to the organisation’s quality system, environmental conditions and application consequence still matter, and any contractual or regulatory requirement still takes precedence. One or two certificates do not justify an extension on their own.
Calibrate the measurement, not the label on the enclosure
| Measurement | What to consider between calibrations | What can influence interval |
|---|---|---|
| Wind | Status, comparison, physical condition | Application consequence, environment, sensor technology |
| Humidity | Plausibility and contamination | Element drift, pollution, dust |
| Temperature | Comparison and shield condition | Siting, radiation, environment |
| Pressure | Comparison against a trusted reference | Application requirement, sensor behaviour |
| Rainfall, tipping bucket | Mechanical inspection and blockage | Mechanical condition, contamination |
| Rainfall, optical | Optical-path condition and verification | Contamination, application requirement |
The same weather station can justify different calibration strategies
Environmental monitoring. A long-running dataset may place greater emphasis on continuity, traceability and consistency over time.
Operational safety. Where the measurement influences an immediate operational decision, the consequence of error may justify stronger assurance.
General site monitoring. Where the data is informational rather than safety-critical or regulated, the required assurance may be different.
Calibration is part of measurement assurance, not the whole of it
A calibration certificate tells you how an instrument compared with a reference under defined conditions. The measurement system also depends on where the sensor is installed, how it is configured, how it is connected, what the environment does to it, how the data is handled and how changes are detected between formal calibration events.
A good calibration strategy therefore has two parts: periodic evidence from controlled comparison, and ongoing evidence from the installation itself.
The strongest calibration interval is one you can explain, evidence and defend.
The short version
- There is no universal calibration interval for a weather station.
- Start with the requirement and the consequence of measurement error, then consider environment and sensor behaviour.
- Different weather-station parameters age and fail differently.
- Calibration, adjustment, verification and production testing are not interchangeable.
- Use historical results and ongoing checks to make the interval defensible, where the governing requirements allow it.
Frequently asked questions
There is no universal calibration interval for every weather station. The appropriate interval depends on the measurement requirement, consequence of error, operating environment, sensing technology, organisational quality requirements and any regulatory, contractual or customer requirement.
Not necessarily. Annual calibration may be appropriate for a particular application, but it should not be treated as a universal rule. If no external requirement specifies an interval, the organisation should establish one based on risk, environment, sensor behaviour and previous calibration results.
The main factors are the purpose of the measurement, consequence of an incorrect reading, operating environment, sensing technology, quality requirements and historical calibration results. External standards, contracts or regulators may also specify the interval.
It can be where the governing requirements permit it and there is sufficient evidence to support the decision. Repeated calibration results that remain comfortably within tolerance can provide evidence for reviewing an internally controlled interval, but the decision must also consider application consequence and environmental conditions.
An interval may need review when calibration results approach or exceed tolerance, the environment becomes harsher, contamination increases, sensor behaviour changes, application consequences increase or quality requirements change. A marginal result should prompt investigation rather than automatically dictating a new interval.
No. A weather station can contain several different sensing technologies, and they do not necessarily drift or fail in the same way. Wind, humidity, temperature, pressure and rainfall should therefore be considered individually when setting an appropriate calibration strategy.
Relative humidity is particularly susceptible to drift and contamination, especially in polluted or dusty environments. That does not mean humidity always determines the calibration interval, because the application and governing requirements still have to be considered.
Ultrasonic wind sensors do not have the mechanical wear mechanism associated with rotating anemometers. Their fixed geometry and timing-based measurement create a different assurance question from a mechanical sensor. Ongoing verification and the manufacturer's production-control method remain relevant.
The need for formal calibration depends on the application and its quality requirements. Some ultrasonic sensors are produced and verified using controlled manufacturing and wind-tunnel processes, while a customer may still require periodic calibration or verification for traceability or quality assurance.
A calibration certificate provides evidence of how the instrument compared with a reference under the conditions covered by the calibration. It does not automatically prove correct installation, siting, configuration, data transmission, application suitability or future performance.
Calibration compares an instrument's indication with a reference under defined conditions and documents the result. Verification checks whether the instrument remains within an expected specification without necessarily changing its response. The two activities provide different forms of measurement assurance.
Calibration determines how an instrument compares with a reference. Adjustment changes the instrument's response to bring it towards the required relationship with the reference. An instrument can therefore be calibrated without being adjusted.
Production verification checks that instruments produced through a manufacturing process remain within the required specification. It is different from individually calibrating or adjusting every instrument.
Gill uses a GenCal production-verification process in which every fifth unit is taken to a highly accurate wind tunnel to check that production remains within specification. This is a production-control process and is not individual calibration of every unit.
No. Gill uses different manufacturing and verification approaches depending on the product. Some products have accuracy established through controlled manufacturing, transducer accuracy and matching, while higher-value products such as WindMaster are individually adjusted during manufacture and every unit is tested in the wind tunnel.
Yes. Higher-value products such as WindMaster are individually adjusted during manufacture and every unit goes through wind-tunnel testing.
No. Different products use different approaches to establishing and verifying accuracy, and some products cannot be individually calibrated or adjusted in the conventional sense. The appropriate assurance method depends on the product and application.
GenCal is Gill's production-verification process for wind sensors. Every fifth unit is taken to a highly accurate wind tunnel to check that the production process remains within specification.
No. GenCal is production verification rather than individual calibration. Its purpose is to check that the manufacturing process remains within specification.
No. UKAS accreditation applies to the laboratory and the activities within its accredited scope. It does not mean the instrument itself is UKAS accredited. The certificate provides evidence of calibration performed under the laboratory's accredited scope where applicable.
Labcal is a UKAS-accredited calibration laboratory within Gill Group. It is UKAS Lab 0625, operating to ISO/IEC 17025:2017, with an accredited scope covering air velocity, temperature and humidity, pressure, gas flow and liquid flow.
Yes. Labcal supports equipment from other manufacturers as well as Gill equipment, within its applicable calibration capabilities and scope.
No. Labcal's capabilities are defined by its accredited scope and the equipment that can be calibrated within that scope. Not every Gill product is necessarily individually calibratable or covered by every laboratory capability.
There is no universal interval. Relative humidity elements can be more susceptible to drift and contamination than some other weather-station parameters, so the application, environment, quality requirements and previous calibration results should determine the interval.
The appropriate interval depends on the application, required accuracy, environment and quality requirements. Temperature sensors are generally stable, but installation and radiation-shield conditions can affect the measurement, so the physical installation should also be checked.
There is no universal interval. Pressure sensors are generally stable and can often be checked between formal calibration events by comparison with a trusted reference, but the formal interval should still follow the application and applicable quality requirements.
The appropriate interval depends on the rainfall technology, application and quality requirements. A tipping-bucket gauge contains a mechanical measurement mechanism and can be affected by blockage, while an optical gauge has no moving mechanical element and instead requires attention to verification and keeping the optical path clean.
You can perform useful verification and plausibility checks between formal calibrations, such as comparing suitable parameters against a trusted nearby reference, checking instrument status and inspecting the installation. These checks do not replace formal calibration where calibration is required.
Where access is expensive or disruptive, it can be practical to combine calibration with planned maintenance or other site work. The aim is to reduce unnecessary visits while maintaining the required measurement assurance.
Retain the calibration certificate, instrument identity, configuration information and relevant service or installation records. For long-running datasets, also record when the instrument was removed, replaced and returned to service so the measurement history remains understandable.
Investigate the result and determine whether the instrument, environment, installation or measurement process has changed. Depending on the cause and the application's requirements, the instrument may require adjustment, repair, replacement, a shorter calibration interval or additional verification.
No. Laboratory calibration checks the instrument under the defined calibration conditions. Installation, siting, orientation, environmental exposure, configuration and data transmission require separate checks as part of the overall measurement system.
Keeping the configuration helps a replacement instrument reproduce the intended measurement setup. This can support continuity when an instrument is removed for calibration, repair or replacement.
Use the governing requirement, measurement purpose, consequence of error, environmental exposure, sensing technology and historical calibration results. Also consider how the measurement is monitored between formal calibration events.
The questions that usually follow
How are wind sensors calibrated?
What happens in the wind tunnel, what the certificate records, and how traceability is established for a wind measurement.
Read the answerSpecifying and choosingWhat should I consider when choosing a weather station?
Parameters, siting, outputs and maintenance decide a weather station choice. A practical route from the measurement requirement to a specific model.
Read the answerMeasurement technologyOptical rain gauge vs tipping bucket
Two rainfall technologies with different failure modes, and what each one needs between calibration events.
Read the answerHave an application to discuss?
Speak to our team
Tell us what you need to measure, where and under what conditions. We'll help you work out the appropriate approach.