The purpose of this document is to provide high-level guidance to National Meteorological Hydrological Services (NHMSs) and network managers in planning a successful transition from mercury-based instruments and other obsolete instruments to modern alternatives.
The main driver for replacing mercury-based instruments is clearly the Minamata Convention on Mercury:
“The Minamata Convention on Mercury of the United Nations Environment Programme (UNEP) came into force globally in August 2017, and bans all production, import and export of observing instruments (thermometers, barometers, etc.) containing mercury (UNEP, 2017). This agreement is a global treaty to eliminate the use of mercury to protect both human health and the environment from the adverse effects of mercury. As a result, national meteorological organisations must transition away from mercury-based instruments. For most countries this will lead to the replacement of conventional instruments containing mercury with electronic ones.” (WMO-№8)
In addition to the Minamata Convention, emerging technologies, as well as ever increasing requirements for environmental data, are providing increased opportunities and requirements for NMHSs to upgrade and expand surface-based meteorological observation networks which reflect developing user needs.
Automatic Weather Stations (AWS) are playing an increasing role in meteorological observing networks in both developed and developing countries. They offer numerous advantages in weather applications by allowing observations at a high time resolution to be received in real time. Many locations that previously had observations only a few times per day are now able have a continuous feed of data every minute. AWS also allow observations to be made in areas without permanent human populations.
On the other hand, AWS typically require more frequent and more specialized maintenance than manual systems and there are the communication costs in transmitting the data where telemetry is used. These activities and costs place a significant burden on network managers, especially in countries with limited resources.
The transition to automation also introduces challenges for long term climate records (WMO-№1202). For example, different response times between traditional and modern instruments may cause inhomogeneities in the maximum and minimum temperature record.
The meteorological and metrological aspects covered in this guidance should help to ensure homogeneity of data series and sustainability of measurements.
As such, an automation programme is a major undertaking for a NMHS and should be managed in a project framework. The Education and Training Office at WMO have provided guidance in A Compendium of Topics to Support Management Development in National Meteorological Services (ETR-24). See sections 10, 11 and 12 in particular.
It is essential that any transition to automated observations needs to consider the complete data chain from observations, data processing and data management to information systems, services and products, as well as related human resource aspects.
Therefore, an appropriate solution depends on many factors including stakeholder and user requirements, environmental conditions, site security, competencies and financial capability. The designed network should be modular enough to be able to accept new hardware and software upgrades. Other considerations include power supply requirements, communication platforms, grounding, lightning protection, and access to the station.
While there are various alternatives to meet the needs of different NMHS with suitable levels of accuracy and ease of use, there are no maintenance-free alternatives. The level of maintenance required may depend on the environmental conditions. Operational and life cycle budgeting, as well as staff training are vital to the sustainability and success of any observation programme and more so in any transition to automation.
A successful transition to automation from conventional mercury based or obsolete instruments entails:
1. Involving stakeholders in the planning and implementation of the phase-out of dangerous and obsolete instruments;
2. Conducting or updating an inventory to determine the quantity and location of dangerous and obsolete devices;
3. Choosing the alternative solution, validate it with stakeholders and have it approved by management;
4. Developing a budget for replacement within a phase-in schedule including a period of parallel measurements by the both old and new systems. At climatological stations this period should be for at least two years;
5. Conducting training activities related to the phase–out of dangerous and obsolete instruments and the phase-in of the alternative solution;
6. Implementing procedures of storage and clean-up of mercury containing devices and transferring dangerous instruments to an appropriate storage/disposal area and remove mercury in accordance with the local regulations on hazardous waste.
Transitioning from mercury-based and other obsolete or unserviceable meteorological instruments is a complex process and a big challenge is in knowing what to do.
The following sections in this guidance note are intended to give a brief but concise summary of options and considerations in transitioning to automation with the annexes providing more detail on different technologies and listing earlier WMO publications and conference proceedings on automation to assist network managers in deciding on the best option for their circumstances.
List of related WMO publications
This guidance note should be read in conjunction with other WMO publications including:
I. A Compendium of Topics to Support Management Development in National Meteorological Services (ETR-24)
II. Guide to Meteorological Instruments and Methods of Observation (WMO-№8, also referred to as the (CIMO Guide));
III. Guide to Climatological Practices (WMO-№100);
IV. Guidance on Automatic Weather Systems and Their Implementation (WMO-№862);
V. Manual on the WMO Integrated Global Observing System (WMO-№1160);
VI. Guide to the WMO Integrated Global Observing System (WMO-№1165);
VII. Guidelines on the Role, Operation and Management of National Meteorological and Hydrological Services (WMO-№1195);
VIII. Challenges in the Transition from Conventional to Automatic Meteorological Observing Networks for Long-term Climate Records (WMO-№1202);
IX. Compendium of WMO Competency Frameworks (WMO-№1209)
The Minamata Convention on Mercury is the major driver in the transition to automation with the phasing out of mercury products in 2020.
In addition, manufacturers are no longer producing outdated traditional instruments. It is therefore inevitable that NMHSs need to consider options for transitioning to modern alternatives. Table 1 lists instruments that contain mercury and must be replaced, and others that are considered obsolete or unserviceable and should be considered for replacement by modern alternatives. List of instruments containing mercury or considered obsolete or unserviceable are:
Temperature: Mercury-in-glass thermometers; Liquid-in-glass thermometers; Thermographs; Mercury-in-steel soil thermographs
Humidity: Mercury-in-glass psychrometers; Hygrographs
Atmospheric Pressure: Mercury barometers; Barograph
Surface Wind: Anemographs; Cup counter anemometer
Precipitation: Chart recording gauges
Sunshine Duration: Campbell-Stokes sunshine recorder
Evaporation: Pan evaporimeter
Sensor and instrument selection
The need to replace mercury-based instruments has been well documented and is well understood. Other reasons to replace instrumentation include:
· Inability to maintain the instrument due to lack of replacement parts;
· Inability to calibrate the instrument;
· Difficulties meeting WMO specification requirements.
· New cost-effective measurement technique.
Instruments and systems should be carefully chosen and the entire chain (sensor, instrument, measurement, exposure, observation, data transfer, data archiving, end users) must be considered and the independent parts must be compatible.
The following should be considered when selecting an instrument or transitioning to an AWS:
· Technical specifications of the instrument, such as the response time of the sensor, stability and reliability, long-term drift. The instrument’s performance specifications should comply with information in the WMO №8, Volume I, Annex 1.A, “Operational measurement uncertainty requirements and instrument performance” for the variables most commonly used in synoptic, aviation and marine meteorology, and climatology;
· The technical specification must be suitable for the intended use and location. For example, there is little point in stipulating a temperature range of -80 to +60°C as per WMO №8, Volume I, Annex 1.A. when the minimum temperature ever recorded in a country is 5°C. In this case, a lesser range of -40 to +60°C used in some common systems is equally acceptable;
· The difference between sensor specification and overall measurement uncertainty. For example, the response time of a temperature sensing element will be affected by the protective probe and type of radiation shield used;
· The suitability of an instrument to be used in harsh environments or extreme weather conditions. Is the instrument suitable for use in extreme heat and cold? For example, will the instrument display last in tropical locations? See WMO-№8, Volume I, Annex 1.F, “Operating equipment in extreme environments”;
· Laboratory and field evaluations of differing potential systems;
· Installation. Who is responsible for installing the stations? Is this an opportunity for training?
· Ease of maintenance, servicing and calibration. Are spares and spare parts readily available?
· Siting it important and the Siting Classification scheme should be used. Many new systems have a small footprint and network managers need to guard against making compromises in siting because it is the easy option.
· Changes in technology and methods can cause inhomogeneities in data series. See Challenges in the Transition from Conventional to Automatic Meteorological Observing Networks for Long-term Climate Records (WMO-№1202) and WMO-№8. For example, radiation shields used for temperature and humidity measurements come with many different designs and performance capabilities. Any changes should be tested by the user first to check their compatibility with the existing system. Suitable overlap periods can help to avoid inhomogeneous data series;
· Not all new instruments and emerging technologies meet reliability or specification standards required by NMHSs. Other factors may also contribute to their incompatibility, such as improper instrument exposure;
· Support of manufacturer
· How will the data get into the national archive? Some proprietary systems have “locked in” data transfer systems making them very difficult for data to be archived in the national archive.
· Does the NMHS have the necessary technical resources (including staff competence) and infrastructure to maintain an AWS?
The decision-making process must be performed carefully to address the differences in instrumentation type, measurement methods, data processing, data control, communication interface and issues related to calibration and maintenance of the instrument.
Temperature
The choice of thermometer type to be used in meteorological applications is based on the observation requirements.
A variety of thermometers are available on the market using various technologies and with different specifications for use in differing applications. Those listed are deemed to have an uncertainty equal to or better than mercury-in-glass thermometers for use in meteorological applications.
· Platinum resistance thermometer
· Thermistor
· Alcohol-in glass thermometer
The most common instrument used in AWS is a platinum resistance thermometer due to its many excellent metrological properties. Thermistors can also be acceptable but are less common. For a manual operation, alcohol in glass thermometers are suitable for most applications but are not suitable for maximum temperature.
Other types include Thermocouple, IC Sensor, and Bimetallic but these have some disadvantages and are less commonly used.
Humidity
Traditionally relative humidity measurements have been made using the dry and wet bulb psychrometer method using mercury-in-glass thermometers. Nowadays, the most widely used humidity instruments, especially in automated weather stations, are based on the electrical capacitance type sensors.
The other common alternative is the psychrometric method using two platinum resistance thermometers in a dry and wet bulb configuration.
Other alternatives such as electrical resistance sensors can also be acceptable in certain situations but are less common and are not considered here due to their limited range of 15% — 95%. Alcohol-in-glass psychrometers are also not considered as they typically have a slower response than mercury.
Atmosperic Pressure
Most modern pressure sensors operate on the principle of converting a pressure change into a mechanical displacement or deformation. Deformation of the sensing element may be converted into an electrical signal that is processed by the measuring system.
There are many types of sensors available and can be used for the measurement of atmospheric pressure. The more common types used in meteorological applications are:
· Variable capacitance (Capacitive)
· Mechanical pressure sensors
Barometers using variable capacitance (Capacitive) sensors are widely used in automated weather stations as they are generally stable, linear and temperature compensated
For manual operations, capacitive or mechanical pressure sensors such as in aneroid barometers are more common.
Other alternatives can also be acceptable but are less common. These are the
· Piezo-resistive (Strain Gauge)
· Piezo-electric
· Resonant
· Optical
· Electromagnetic
· Variable Reluctance
· Potentiometric
Due to the wide range of technologies available, pressure instruments vary considerably in their design, performance, and cost with each technology having its own benefits.
Surface Wind
Wind speed and direction can be measured with a variety of instruments, any of which will generally be mounted 10 m above ground level.
The types used in meteorological applications are:
· Cup anemometer and vane
· Ultrasonic
· Acoustic resonance
· Propeller vane
Ultrasonic wind systems (including resonance type) are replacing mechanic cup and vane anemometers and propeller vane anemometers as the most common types in use due to their perceived low maintenance requirements but all types are suitable for use on AWS.
All anemometer types require built-in heating systems when used in a cold environment, to prevent freezing. Anemometer heating systems significantly increase power consumption.
The use of acoustic resonance sensors is useful in extreme environments and in mobile applications where the robustness, portability and small size of the sensor are important.
Care must be taken to ensure that the wind field deformation caused by the transducers is corrected appropriately by the sensor.
Sometimes ultrasonics can be affected/damaged by birds.
Precipitation
The most common instrument used in an automated meteorological observation system is the tipping bucket precipitation gauge. Other options such as weighing gauges are also becoming more common especially in areas that experience solid precipitation.
For manual use the conventional storage type gauge, also referred to as an ordinary gauge, and mechanical chart recorders are still suitable.
The important requirements of precipitation gauges are:
· Sharp rim, vertical on inside and steeply bevelled on outside;
· Area of orifice is known to nearest 0.5% and construction should be robust enough this does not change over time;
· The collector is designed to prevent rain splashing in or out;
· Construction and materials used such as to minimize wetting errors;
· The container should have narrow entrance and protected from radiation to minimize loss of water by evaporation.
New gauges can be different in size/height to older gauges. Users need to be aware that the introduction of a new sensor may undercatch due to wind deformation. Heating might be required where solid precipitation is expected.
Also one moves from (daily) precipitation totals towards precipitation intensity and duration.
Sunshine duration
Various principles can be employed for measuring sunshine duration: these include pyrheliometric, pyranometric, burn, and contrast-evaluation.
The most common and simplest method used on AWS is the pyranometric and contrast-evaluation methods.
The pyrheliometric method is also suitable but much more expensive and requires power. This method is more suitable for more demanding applications.
The burn method is only suitable for manual operations
